The Political Economy of Craft Production: Crafting Empire in South India, c. 1350-1650

This page intentionally left blank The Political Economy of Craft Production Crafting Empire in South India, c. 1350–...

Author: Carla M. Sinopoli

169 downloads 1186 Views 7MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

This page intentionally left blank

The Political Economy of Craft Production Crafting Empire in South India, c. 1350–1650

The study of specialized craft production has a long tradition in archaeological research. Through analyses of material remains and the contexts of their production and use, archaeologists can examine the organization of craft production and the economic and political status of craft producers. This new study combines archaeological and historical evidence from the author’s twenty years of fieldwork at the imperial capital of Vijayanagara to explore the role and significance of craft production in the cities’ political economy of the fourteenth to the seventeenth century. By examining a diverse range of crafts from poetry to pottery, Sinopoli evaluates models of craft production and expands upon theoretical and historical understandings of empires in general and Vijayanagara in particular. It is the most broad-ranging study of craft production in South Asia, or in any other early state empire. carla m. sinopoli is Professor in the Department of Anthropology at the University of Michigan and Curator of Asian Archaeology in the Museum of Anthropology at the University of Michigan. She is the author of Pots and Palaces: The Earthenware Ceramics of the Noblemen’s Quarter of Vijayanagara (1993), Approaches to Archaeological Ceramics (1991) and co-editor of Empires: Comparative Perspectives from Archaeology and History (2001).

The Political Economy of Craft Production Crafting Empire in South India, c. 1350–1650

carla m. sinopoli University of Michigan

   Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo Cambridge University Press The Edinburgh Building, Cambridge  , United Kingdom Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521826136 © Carla M. Sinopoli, 2003 This book is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published in print format 2003 - isbn-13 978-0-511-07101-0 eBook (EBL) - isbn-10 0-511-07101-9 eBook (EBL) - isbn-13 978-0-521-82613-6 hardback - isbn-10 0-521-82613-6 hardback

Cambridge University Press has no responsibility for the persistence or accuracy of s for external or third-party internet websites referred to in this book, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

In memory of Richard Carl Sinopoli and Channabasappa S. Patil

Contents

List of figures [page viii] List of tables [xi] Acknowledgments [xii] 1 Introduction: crafting empire in South India [1] 2 Specialized craft production: archaeological approaches [13] 3 The South Asian state [38] 4 Vijayanagara: the historical setting [63] 5 Vijayanagara: sources of evidence [119] 6 Craft products and craft producers [156] 7 Artisans and institutions: artisans and each other [252] 8 Crafting empire: conclusions [295] References [317] Index [342]

Figures

4.1 Core areas of pre-Vijayanagara states of South India [page 67] 4.2 Vijayanagara boar emblem from gate at Chandragiri (photo by the Vijayanagara Metropolitan Survey Project) [71] 4.3 The Deccani sultanates and Vijayanagara empire [77] 4.4 The Vijayanagara empire: upland core [84] 4.5 The Vijayanagara region: the Tungabhadra River valley (photo by the Vijayanagara Metropolitan Survey Project) [85] 4.6 Vijayanagara-period seaports and major overland routes [86] 4.7 Jalakantesvhara Temple, Vellore (photo by the Vijayanagara Metropolitan Survey Project) [95] 5.1 Areas included in inscriptional database [123] 5.2 Iron smelting and forging facilities depicted by Frances Buchanan [137] 5.3 The Vijayanagara urban core [143] 5.4 The Tiruvengalanatha temple (photo by the Vijayanagara Metropolitan Survey Project) [144] 5.5 The Vijayanagara metropolitan region [147] 5.6 Site distribution in the intensive survey area of the Vijayanagara metropolitan region [149] 5.7 Varadadevi-amana-pattana: distribution of archaeological remains [151] 5.8 Vijayanagara’s defensive infrastructure [153] 5.9 Vaishnava temple complex (VMS-142), Varadadevi-amana-pattana (photo by the Vijayanagara Metropolitan Survey Project) [154] 6.1 Female performers on the basement of mandapa in the Vitthala temple complex, Vijayanagara Sacred Center [168] 6.2 Bearded male performers on royal throne platform, Vijayanagara urban core [169] 6.3 Female performers sculpted on royal throne platform, Vijayanagara urban core [170] 6.4 Centers of textile production in Vijayanagara-period South India [175]

List of figures

6.5

6.6 6.7

6.8 6.9 6.10 6.11

6.12

6.13

6.14 6.15

6.16 6.17

6.18

6.19 6.20 6.21

Lepakshi painting: group of women attending the goddess Parvati (published with permission of the American Institute of Indian Studies) [181] Lepakshi painting: group of women (published with permission of the American Institute of Indian Studies) [182] Sculpted figure on gate in outer city walls depicting pointed kullayi headgear (photo by Vijayanagara Metropolitan Survey Project) [184] Iron-working sites in the Vijayanagara metropolitan region [198] VMS-121: Iron-working site and shrine (photo by Vijayanagara Metropolitan Survey Project) [199] Hilltop fort and associated settlement (VMS-6) [200] Administrative architecture in the Vijayanagara urban core: royal elephant stables (photo by Vijayanagara Metropolitan Survey Project) [210] Dry-stone masonry in the Vijayanagara urban core: detail of wall construction (photo by Vijayanagara Metropolitan Survey Project) [212] Section of fortification wall in the Vijayanagara urban core: trapezoidal blocks (photo by Vijayanagara Metropolitan Survey Project) [212] Fortification wall (VMS-339), constructed of unmodified boulders [214] Quarry marks incised on large outcropping boulder in Vijayanagara metropolitan region (photo by Vijayanagara Metropolitan Survey Project) [215] Quarrying site in the Vijayanagara metropolitan region (VMS-341) (photo by Vijayanagara Metropolitan Survey Project) [216] Uncompleted structure (VMS-79): blocked out column and trimmed blocks (photo by Vijayanagara Metropolitan Survey Project) [217] Truncated columns in the Vitthala temple complex, Vijayanagara urban core (photo by Vijayanagara Metropolitan Survey Project) [219] Carving on the walls of Raichur Fort depicting transport of large quarried column [220] Finely sculpted Hanuman image on freestanding slab (VMS-202; (photo by Vijayanagara Metropolitan Survey Project) [223] Finely sculpted goddess image on outcrop (photo by Vijayanagara Metropolitan Survey Project) [224]

ix

x

List of figures

6.22 Crudely carved depiction of Rama and Sita images on boulder in Vijayanagara metropolitan region (photo by Vijayanagara Metropolitan Survey Project) [224] 6.23 VMS-538: unfinished Hanuman image [225] 6.24 Seated royal figure sculpted on the face of the royal throne platform, Vijayanagara urban core (photo by Vijayanagara Metropolitan Survey Project) [227] 6.25 Memorial stone from Bhatkal, Western India (photo by Vijayanagara Metropolitan Survey Project) [229] 6.26 Memorial stone of warrior on horseback in the Vijayanagara metropolitan region (photo by Vijayanagara Metropolitan Survey Project) [230] 6.27 Memorial stone dedicated to a sati, Vijayanagara metropolitan region (photo by Vijayanagara Metropolitan Survey Project) [231] 6.28 Illegible inscription at small temple complex in Vijayanagara metropolitan region (photo by Vijayanagara Metropolitan Survey Project) [235] 6.29 Completed and uncompleted bedrock mortar (VMS-101) (photo by Vijayanagara Metropolitan Survey Project) [237] 6.30 Vijayanagara ceramics: vessel forms [242] 6.31 Contemporary pottery production: paddle and anvil work [244] 7.1 Inscriptions referring to Kammalar and Kaikkolar in the Vijayanagara administrative district Valudilampattu-rajyam [268] 7.2 Valangai-idangai uprising inscription locations [287]

Tables

2.1 Typologies of the organization of craft production [page 19] 5.1 Inscription types [126] 5.2 Foreign travelers in fourteenth- through seventeenth-century South India [127] 5.3 Selected Vijayanagara-period literary works [131] 5.4 Summary of South Canara census data, 1807 (craft producers in bold) [138] 5.5 Primary site functions documented by the Vijayanagara Metropolitan Survey [150] 6.1 Kaikkolar weavers in Tirupati-Tirumalai inscriptions [189] 6.2 Iron-working locales in the Vijayanagara metropolitan region [197] 6.3 Peg production in the Vijayanagara metropolitan region [236] 7.1 Categories of administrators able to reallocate taxes [265] 7.2 Craft producers as recipients of shares from temple donations at Tirupati-Tirumalai temples [274]

Acknowledgments

In the spring of 1981, when I was studying for my doctoral exams in Near Eastern archaeology, I received a letter from John Fritz, inviting me to join an archaeological field project he was beginning at the site of the historic South Indian capital of Vijayanagara. His letter started a journey that has lasted for more than twenty years, and if this book is not its culmination, it is at least an important way point. My first fieldwork at Vijayanagara took place in 1983 and 1984 as part of my doctoral research on the earthenware ceramics from the city. In 1988, I began a new field project, in collaboration with Kathleen D. Morrison – The Vijayanagara Metropolitan Survey – the field component of which ended in 1997. Over the last twenty-plus years, I have accumulated many debts, which I gratefully acknowledge here. I extend my deepest appreciation to the Government of India and Archaeological Survey of India (ASI) for granting approvals to me to conduct this research. I also thank the ASI regional staff and directors at Vijayanagara for their help. Throughout my work, I have been affiliated with the Karnataka Department of Archaeology and Museums, whose staff members have been sponsors, friends, and collaborators. I thank the Department’s directors and particularly retired director Dr. M.S. Nagaraja Rao, who welcomed me to Vijayanagara in 1983 and has been supportive of my work ever since. I also thank the Department staff including T.S. Gangadhar, Balasubrahmanyam, and Manjunataiah. My particular gratitude goes to Channabasappa S. Patil, who was a doctoral student working for the department when I arrived in 1983, and later its Deputy Director, until he passed away unexpectedly in 2001. We grew up together in South Indian archaeology. Dr. Patil was a fine scholar, a dedicated fieldworker, and a good and generous friend. He is deeply missed. The American Institute of Indian Studies has been a source of support and guidance throughout my years at Vijayanagara. I have had the pleasure of being affiliated with the Institute since my first season in India, and I am extremely grateful for all of the help provided by their dedicated staff in Delhi and Chicago. I owe especial thanks to Dr. Pradeep Mehendiratta; it is no exaggeration to say that this work could not have taken place without his guidance.

Acknowledgments

My thanks also to the many funding agencies that have supported our fieldwork over the years. These include the American Institute of Indian Studies, Asian Cultural Council, National Geographic Society, National Endowment for the Humanities, National Science Foundation, Smithsonian Institution, and Wenner Gren Foundation for Anthropological Research. The University of Wisconsin-Milwaukee and University of Michigan also provided much appreciated support. Support for writing this book came from a National Endowment for the Humanities Fellowship; thanks also to the Department of Anthropology at Arizona State University for allowing me to spend my sabbatical year in Tempe working on this manuscript. John Fritz has been a valued mentor, friend, and colleague for more than twenty years. I thank him for inviting me to Vijayanagara and for all of his support over the years. Thanks also to George Michell, whose friendship and wisdom I also treasure, as well as to the many other scholars who have passed through Vijayanagara. I extend my deepest appreciation to my co-director in the Vijayanagara Metropolitan Survey, Kathleen D. Morrison. When we decided to direct a small survey around Vijayanagara, we had no idea that we were launching a collaboration that would last for more than fifteen years (and still counting!). Throughout Kathy has been a source of friendship, support, and a never-ending supply of interesting ideas. During the course of our survey, we were assisted by dozens of field assistants and co-workers. I thank the many graduate students from Deccan College, Pune who joined us in the field, as well as Dr. V.S. Gogte from the same institution. Thanks also to the many American students who joined us from the Universities of Pennsylvania, Michigan, Chicago, Wisconsin, Hawaii, and Northwestern University. Of our many project members, I single out three for special mention. Mark Lycett worked with us for several seasons in the field and has been a source of advice and ideas and a friend throughout. Rob Brubaker also joined us in the field, and developed a doctoral project to examine Vijayanagara’s defensive infrastructure; he also designed the project database and geographic information system. U.V. Srinivas worked with us through the entire project, as our driver and master of all trades; we could not have accomplished much of what we did without his ready and creative assistance. Several people read and provided valuable comments on all or parts of this manuscript. I thank the reviewers for Cambridge University Press, Cathy Costin, Mark Kenoyer, and one anonymous reader, for their thoughtful comments and suggestions. Kathy Morrison and George Michell each read the complete manuscript and provided valuable guidance. Norman Yoffee read chapter 3 and provided valuable insights and references. Thanks also to my writing ‘support groups’ at Arizona State University and the University

xiii

xiv

Acknowledgments

of Michigan: respectively, Kate Spielmann, Peggy Nelson and Michelle Hegmon, and Janet Richards and Lisa Young. Many other names belong here, but twenty years of work and life make the list too long to recount and my memory too fragile to be certain that no one is omitted. So, I end by thanking my colleagues, friends, and family for being there in my Vijayanagara and extra-Vijayanagara life. I dedicate this book to the memory of two people who are no longer physically present in my life, but who will never be gone: Channabasappa S. Patil, and my twin brother, Richard C. Sinopoli.

1

Introduction: crafting empire in South India

Specialized craft production and craft producers have a prominent place in archaeological studies of early states and empires. Social and economic differentiation are defining characteristics of such societies, and through analyses of material remains and the contexts of their production and consumption, archaeologists can examine both the organization of production and the social, economic, and political statuses and inter-relations of producers and consumers of craft goods. In this work, I examine the social and political significance of craft production and consumption in the Vijayanagara empire, an expansive polity that dominated much of South India from the fourteenth through seventeenth centuries AD. My study of the political economy of specialized craft production in historic South India situates both political economy and specialized production in the broadest possible frame. I view “political economy” as the relations between political structures and systems (including the constitution of political authority) and the economic realms of production, consumption and exchange (e.g., G. Stein 2001: 356). “Specialized craft production” is understood as the investment of labor by (more or less) skilled practitioners in the production of diverse goods that are in turn consumed by nonproducers. My goals are both to learn more about Vijayanagara and the lives and products of the diverse subjects of this large and complex empire, and to contribute to broader theoretical understandings of empires, imperial economies, specialized production, and archaeological and historical approaches to the study of states in South Asia and beyond. The diverse archaeological and written sources of evidence available on the Vijayanagara period provide rich evidence with which to explore these issues. More than twenty years of systematic archaeological research in the core and hinterland of the first Vijayanagara capital, described in chapter 5, have provided detailed information on a range of material goods produced by specialists, from architecture to earthenware ceramics. These goods defined and constrained the spaces and settings in which political, religious, military, economic, and other activities of daily life occurred in this seminal region of the empire. Excluding temples, which have been documented throughout the empire, only the Vijayanagara capital is known from systematic

2

The Political Economy of Craft Production

archaeological and architectural research. However, the many written sources from the period also provide valuable insights into the social, economic, and organizational structures of craft production, and on many durable and non-durable craft goods. These sources, summarized in chapter 5, include thousands of inscriptions carved on temple walls and copper plates, contemporary South Indian literary works, and the accounts of foreign visitors to the empire, as well as post-Vijayanagara colonial documents. Unlike the available archaeological evidence, the written sources both span the entire empire and extend our temporal range beyond the abandonment of the first Vijayanagara capital in AD 1565. I approach the study of Vijayanagara craft production from two seemingly disparate theoretical perspectives, one drawn from anthropology and the other from South Asian historiography. From anthropology, I build upon the theoretical and comparative literature on the economic, political, and social significances of specialized craft production and material culture. For decades, archaeologists have studied craft production as a route to understanding the emergence and organization of ancient states and empires. Specialist production of diverse categories of craft goods provides both the prestige items necessary to political elites and the essential commodities that fuel the internal and external economies of early states and empires. Archaeological approaches to craft production, summarized in chapter 2, have documented the scale and organization of production units in particular historical settings, and have sought to identify actors and institutions able to exert control over craft production, craft goods, and the people who produced them. As I will elaborate in chapter 2, most discussions of craft production in state societies take as underlying assumptions that both the scale of production units and institutional ability to control production increases in parallel with political complexity. The Vijayanagara evidence provides an opportunity to evaluate these two assumptions in the context of a very large, very complex imperial polity, and in this case, at least, neither is supported. From South Asian historiography, I approach the study of Vijayanagara craft production and political economy from the perspective of ancient and more recent discussions about the nature of the state in precolonial Asia and, particularly, India. As I discuss in chapter 3, in much South Asian historiography images of tyrannical despots have been interwoven with those of autonomous caste-ridden village republics in curious ways. The result is a view of a timeless, history-less, past, in which technologies, identities, political institutions and responses to them are viewed as stable and, indeed, stagnant.


In recent years, numerous scholars have called these long-standing beliefs into question (e.g., Chattopadhyaya 1994; Inden 1990; Kulke 1995a; Talbot 2001), and have proposed alternate models for understanding precolonial South Asian states (e.g., Indian feudalism, segmentary states, patrimonial states; see chapter 3). Several of these have been applied to Vijayanagara, with varying success (see chapters 3 and 4). Nonetheless, earlier concepts of the Asian state have had remarkable endurance, and underlie many historical and archaeological studies of precolonial South Asia. These same perspectives, I believe, have also played a significant and under-appreciated role in the development of anthropological and archaeological approaches to the “ancient state,” creating a point of intersection between my broader anthropological interests and my South Asian concerns. I will return to these questions in chapter 8. While I position my study of Vijayanagara in the context of general models of craft production and the Asian state, I also situate it in the historic particularities of fourteenth- through seventeenth-century South India. The Vijayanagara period was a time of dramatic changes in South Indian society and economy. These changes include the adoption or appearance of new political and military structures and strategies; the expansion and growth of urban centers; increasing monetization and growth in local and longdistance commerce; dramatic expansions in craft and agricultural production; and population growth and redistribution. As I elaborate in chapter 4, many of these trends began in the centuries preceding the emergence of Vijayanagara. However, they intensified and coalesced in new ways during the Vijayanagara period, with important consequences for craft production and craft products, as well as the political structures of the Vijayanagara empire. An additional characteristic of the Vijayanagara period, relevant to the study of craft production, concerns the many and diverse institutions and actors who played important roles in social, political, religious, and economic institutions and spheres. Power, political and otherwise, was distributed among imperial and regional hereditary elites and administrators, military officers, temple institutions and leaders of diverse religious sects, merchant associations, and various caste and regional organizations. Rulers and state institutions were part of this complex array, but were far from the only, or even necessarily the most important, players. The relations between these diverse figures and institutions were neither uncontested nor stable; competition and factionalism (e.g., Brumfiel and Fox 1994) were widespread, creating contentious and shifting political and economic structures and

3

4


relations that varied over both time and space. Historian David Ludden has described South India between c. AD 1200 and 1700 as a terrain of perpetual movement, where social, political, and economic order only emerged in the context of constant, pervasive conflict and adjustment. Vertical conflicts in the social relations of inequality and hierarchy generated patterns of social subordination. Horizontal conflicts among social groups created schisms, partitions, fragmentations and segregations among groups . . . Religious and political elites in these centers of power envisioned the localities of social reproduction as components of their own domain, as units of an orderly system established by their own moral authority. But these temple and court elites exerted their power in a social world that they did not control: elites engaged local concatenations of power that could not be reduced to regulation by temple and court. In pre-modern India, a vast array of local actors exercised structural powers to transform local society, changing their own material world and terms of their social transactions. (Ludden

1996: 109)

For craft production, one of the consequences of this “constant, pervasive conflict and adjustment” was the presence of numerous potential and competing patrons for craft products and their producers. Particularly desired craft goods were those that served as symbols of status and prestige, including elaborate textiles and the works of court poets, as well as monumental temples and palaces, and military paraphernalia. The presence of multiple consumers and patrons for such goods may have provided at least some craft producers with much greater potential for social and economic mobility than was possible in less differentiated, more linearly hierarchical, state systems. There is considerable evidence for social mobility among various craft-producing communities and individuals throughout the Vijayanagara period (see chapters 6 and 7). Along with being patrons and consumers of craft products, many of the diverse South Indian elites of the Vijayanagara period gained substantial economic benefits from the expansion of craft production and commerce. Taxation on raw materials, production, finished goods, and commerce provided important revenues to political and military leaders. As I discuss in chapter 7, taxes were assessed and collected, usually in currency, at a variety of levels – by kings, imperial administrators, and their representatives; military officers; hereditary local elites; caste and merchant organizations; and village and town councils. Tax payers, predominantly non-elite artisans and agriculturalists, were well aware of the complex and dynamic political conditions under which they lived, and there are several documented cases of their successful resistance of excessive taxation through large-scale collective action. Taxes that were collected were reallocated and deployed


through a variety of hierarchical and horizontal routes. Some, often only a small percentage, flowed from local administrators to imperial coffers. In other cases, individuals authorized to collect taxes reallocated them to religious institutions or to individuals associated with temples. These institutions, many of which grew to considerable size during the Vijayanagara period, both employed large numbers of artisans in a variety of capacities and were actively involved in the economic expansion that characterized the Vijayanagara period, particularly the growth of agricultural and craft (particularly textile) production (see chapters 4 and 7). While taxation and the resultant revenue generation and redistribution were critical to the Vijayanagara political economy, it is important to note that although diverse institutions benefitted from the revenues generated by craft production, there is no evidence that they ever sought to directly regulate craft goods or administer their production. During the Vijayanagara period, the status and organization of individual craft-producing communities differed considerably. These differences were the result of a complex array of factors, including the nature and value of the goods produced, as well as the historical, social, and physical contexts in which production occurred. It is thus not possible to use a single craft to characterize Vijayanagara craft production, and in this study, I examine a very broad array of material and non-material crafts. Before examining the complexity and variation in Vijayanagara craft production and political economy, I begin here with some broad generalizations. First, I have already noted that most crafts and craft producers were not directly administered by the Vijayanagara state or by other institutions, such as temples. This does not mean that certain acts of production and producers were not “attached” to institutions in a variety of ways, but attached specialization was, overall, not the dominant social relation of production for most crafts. Second, Vijayanagara craft production was characterized by a very high degree of economic specialization. This is most dramatic in textile production, where weavers, dyers, washers, fabric painters, and textile merchants each constituted a discrete occupational group (or groups; see chapter 6). Third, most, though certainly not all, craft producers were members of hereditary groups – castes, subcastes, and lineages – each with a unique history, social identity, traditional occupation, and social status (though the latter, at least, was often contested). While membership in hereditary caste groups was an important factor influencing both social identity and occupation, it is also important to note that there was considerable economic,

5

6


social, and occupational mobility during the Vijayanagara period, at both individual and group levels. A fourth, and perhaps obvious, generalization about Vijayanagara craft production is a necessary consequence of the high degree of economic specialization noted above. As a result of this specialization, producers were necessarily interconnected through complex webs of interaction and interdependence. Specialist producers required raw materials, processed goods, and finished artifacts from other artisans, as well as foodstuffs available from agriculturalists or in markets. The relations among diverse artisans were structured and regulated at a variety of levels, involving individual producers and workshops, caste organizations, merchant associations, and, rarely, the state. A fifth, and far from obvious, generalization about Vijayanagara craft production is that despite the complexity and intensity of craft production in fourteenth- through seventeenth-century South India, we do not see the development of large-scale, centrally administered, units of craft production, such as factories or imperial workshops. Even as the demand for craft products increased significantly, the vast majority of craft production activities took place in small-scale household workshops. I discuss a number of exceptions to this later in this work, but the dominance of small-scale household production in conditions of high output and demand and within an extremely complex political landscape is striking, and quite different from what most models of craft production in state societies would lead us to expect. The disparities between the Vijayanagara evidence and many of our expectations for the organization of craft production raise important questions concerning both our theoretical models and the particular conditions of fourteenth- to seventeenth-century South India. In attempting to pursue such questions, this study is situated within contemporary theoretical perspectives in both anthropological archaeology and South Asian history, which explore the complexity and diversity of the kinds of political, economic, and social relations and authority that existed in ancient states and empires, and acknowledge the limitations of centralized control in such contexts. I will return to these themes in the conclusions of this work, after first examining the Vijayanagara evidence in detail. As noted, my study of Vijayanagara employs a diverse range of archaeological and textual sources. The archaeological data are derived from my and others’ fieldwork in and around the eponymous first capital of the empire, and allow an exploration of various craft products and the contexts in which they were produced, consumed, and discarded. Written sources include


inscriptions, literary works, the accounts of foreign visitors to Vijayanagara, and post-Vijayanagara colonial sources. Each of these sources of evidence has inevitable epistemological problems – of translation and interpretation – and each is biased, whether by factors of archaeological preservation, geographic and historical context, or by the political standing and interpretive predispositions of foreign observers or their diverse South Indian creators. I will address these limitations as I present the data. What I do not wish to rehearse is a formulaic discussion of the advantages or disadvantages of historical versus archaeological data. Both are valuable; both are problematic. Nor do I necessarily seek or expect consistency between or among these various sources. Following much contemporary social theory, I expect that complex societies are indeed complex, and often messy – that motivations in one domain may be counteracted by those in others; that people do not necessarily act to enhance the coherence of a systemic whole; and that in imperial contexts in particular the many diverse participants in political and economic relations and transactions may often be in conflict and contradiction with each other; and their decisions and actions may be contingent and responsive to particular situations and events, rather than systemic or systematic. Although seldom neat or consistent, the diverse sources of evidence on Vijayanagara provide windows into the complexities of the period and the lived lives of both the subjects and rulers of empire. I explore these complexities through examining the roles of courts, temples, and diverse social groups and their material products in the production of craft goods and the production of social relations in fourteenth- through seventeenth-century South India. In chapter 2, I present an overview of recent archaeological literature on specialist craft production and its relevance to the study of complex societies. My concern is not with the emergence of specialized production, which has been addressed by Rice (1981, 1991), Arnold (1985), and Costin (1986), among others. Instead, I restrict my focus to those issues relevant to an exploration of the social, political, and economic dynamics of specialized production in the context of a polity characterized by high degrees of economic specialization. Issues addressed include the continuum of attached vs. independent specialization and its relevance to questions of “control,” the archaeological indicators of different productive modes, and questions of agency, identity, and the various kinds of “power” exerted or controlled by producers in early states and empires. Vijayanagara emerged nearly two thousand years after the earliest historic states of South Asia (and four millennia after the Indus Valley states, with their well-documented evidence for productive specialization, but with little

7

8


evidence for any significant impact on later political developments in the region). Within South India, Vijayanagara built on a nearly two millennia long history of social and economic complexity. Specialist production had appeared in the region by the mid-first millennium BC, and artisans and artisan guilds are attested in the literary sources from the early first millennium AD. Although distinctive, and in many respects radically different from the economies of earlier states in the region, the Vijayanagara economy was nonetheless not constructed anew, but built on complex and long-lasting historical developments of statehood, and economic, ideological and social structures, including those of caste differentiation. The inhabitants of preBritish South India were indeed a “people with history” (pace Wolf 1982) and the story of Vijayanagara specialized craft production does not begin from an undifferentiated, unspecialized past. In chapter 2, I propose a very broad definition of “craft.” This definition encompasses goods with material outcomes accessible to archaeology, as well as those with less tangible outcomes that nonetheless involve skilled producers with access to specialized knowledge and techniques/technologies, who manufactured a product for one or more consumers other than, or in addition to, themselves. Thus, I include such archaeologically invisible (or often so) artisans as poets, bards, dancers, and musicians as craft specialists, along with the more conventional weavers, potters, masons, and smiths. Chapter 2 also reviews various archaeological and theoretical approaches to the study of craft specialization. In particular, I focus on models for the organization of specialized production that address the size and composition of productive units, productive scale and intensity, and the relations of artisans to centralized institutions. I next turn to a brief discussion of identity and social action among craft producers. Here, I put the producers as social actors at the center, rather than focusing on producers as “acted upon” by institutions or structural forms, and address the social transmission, contexts, and meanings associated with craft production. I then discuss some recent approaches to material culture, to refocus again, this time on the goods that craft specialists produce and that archaeologists most often rely upon in order to study both specialized production and the broader societies in which goods were made and used. I conclude the chapter with a more explicit discussion of craft and political economy, focusing on relations between production and the state and the diverse roles that craft goods fill in state societies. In chapter 3, I turn to a consideration of the historiography of the Asian state, with a particular focus on South Asia. It may seem unnecessary to


reprieve critiques of Oriental Despotism or the Asiatic Mode of Production in this work. However, I will argue that these views (or various of their intellectual descendants) continue to figure prominently, though seldom explicitly, in interpretations of South Asian states, as well as in how archaeologists have interpreted the region’s prehistoric and historic past, and early states more generally. Images of autonomous isolated villages whose inhabitants plied unchanging technologies and ways of life, and of rulers who are variously portrayed as tyrannical despots or, more commonly today, ritual figureheads, continue to be widespread in both popular media and academic discourse. And in South Asian archaeology, exogenous factors – most often population movement and replacement – continue to dominate accounts of sociopolitical and material culture transformations, and indeed are often considered the only possible causes of change in such innately “static” societies. Many studies of the South Asian state have emphasized the region’s historical uniqueness. While this is of course valid, the failure to contextualize understandings of South Asia in a broader, comparative, intellectual framework is problematic. Judicious attention to appropriate comparative sources can help us to better frame South Asian history, and prehistory, in the context of general understandings of state formation and organization, as well as to examine the distinctive histories of particular cases. Further, by adopting a broader theoretical approach, South Asian specialists will also be in a better position to bring our rich data and interpretations to bear on anthropological theory and approaches. Even today, and despite a wealth of archaeological information and high-quality data, the Indus Valley or Harappan civilization is still frequently portrayed as unknown and unknowable; and few western scholars know even this much about the succeeding periods of state and empire formation in the Ganges Basin, the Deccan, or South India. This book will not address these gaps in archaeological knowledge, but I do wish to briefly explore some of historic underpinnings for their existence. In chapter 4, I turn to the specific historical and archaeological context of my research: the fourteenth- to seventeenth-century South Indian Vijayanagara empire. I summarize recent research on the period, which has been the focus of considerable archaeological and historical scrutiny since the late 1970s. I begin with an historic overview of some of Vijayanagara’s predecessors – the Chola, Kakatiya, Hoysala, and Chalukya states – in order to explore some of the historical context for later Vijayanagara developments. Next, following an outline of dynastic history and imperial geography, I turn to a more detailed focus on several issues of particular relevance to this study:

9

10


political structures and players; the significance of temples; and the nature and constitution of social and economic groups, including occupational communities, village servants, merchant guilds, and caste organizations. I conclude the chapter with a summary of three important perspectives on the Vijayanagara empire – presented by historians K.A. Nilakanta Sastri, Burton Stein, and Noburu Karashima – each with quite different implications for the interpretation of Vijayanagara political economies. In chapter 5, I summarize the nature and range of the textual and archaeological sources of evidence concerning Vijayanagara that provide the primary sources of evidence for this study. I consider their relevance to the study of craft production, as well as their limitations. I briefly describe the recently completed Vijayanagara Metropolitan Survey project, which focused on documenting economic activities in the c. 450-square kilometer hinterland of the imperial capital of Vijayanagara. The geographic biases of this research will be addressed. That is, detailed archaeological research of the estimated 360,000-square kilometer imperial territory has been largely restricted to the c. 450-square kilometer region surrounding the empire’s first capital. This is a minuscule sample of the empire as a whole (studies of Vijayanagara architecture have, however, been far more expansive, see Michell 1995). As noted, text-based historical research has been more extensive, but until quite recently has been biased toward the Tamil-speaking regions of the southeastern area of the empire. In chapters 6 and 7, I focus explicitly on craft production during the Vijayanagara period. I combine archaeological and textual data to address technologies of production, as well as producers as individuals and members of small and large-scale social groups, and in relation to state and religious institutions. Chapter 6 organizes these data by individual crafts or technologies. Some crafts I consider, such as poetry, music, and dance, rarely appear in archaeological studies in concert with such prosaic crafts as ceramic or metal production. However, I include them here both because they are the products of specialist labor, and because they figured importantly in diverse Vijayanagara-period political and ideological arenas. Chapter 6 begins with poetry and ends with potsherds. In between, I discuss musicians and dancers – including the temple women who so intrigued colonial authors – weavers, smiths and metal workers, stone workers, and wood workers. The sources of evidence – written and material – for each of these categories of producers are variable and each craft can be broken down in diverse ways, based on location, caste affiliation, productive technology, or the consumers of their products. For each craft addressed, I attempt to explore the nature and sources of this variability.


In chapter 7, I expand upon the discussion of individual producers and products to examine the relations among craft producers, and between craft producers and the various elites, institutions, and social groupings of fourteenth- through seventeenth-century South India. I focus on three sets of relations: producer–state relations, particularly taxation and tax remissions; producer–temple relations; and relations among producers and between producers and merchants. In this review, I necessarily rely heavily on documentary sources from the Vijayanagara period, including a rare tax document – the Puttasti of Alamkonda – and a database of more than 250 craft-related inscriptions derived from an examination of some 3,000 Vijayanagara-period inscriptions. The latter provide insights into the diverse interactions of craft producers with temples. In addition, several inscriptions also allow examination of local taxation and resistance to it, and legal agreements among producers and between producers and merchants, allowing a consideration of interrelations among diverse producers, including evidence for conflict, as well as corporate social action and inter-regional dynamics. In chapter 8, I attempt to pull together the various strands of this work by summarizing some of the broad themes that have emerged from the study of Vijayanagara craft production. I argue that in imperial states, the presence of diverse economic strategies and dispersed structures of political and economic control, such as we see for Vijayanagara craft production, are likely far more common than many theoretical models of states or craft production have allowed. I review how the particular historical settings through which power and political authority were distributed in fourteenth-century South India influenced the structures of Vijayanagara craft production, and I address several factors that may help explain why large-scale industries and intensive administrative investment in craft production did not develop in this context of high productive intensity and high demand for politically and economically significant craft products. These factors include the abundance of labor and raw materials in many regions of the empire, as well as political circumstances that contributed to the existence of numerous elite patrons for many craft products. Other relevant factors relate to the changing status of craft goods and craft producers in South Indian political economies. Another relevant factor in the Vijayanagara political economy was the presence of powerful and longstanding non-state structures for economic and social regulation. These included merchant associations and durable, hereditary, geographically and occupationally defined social groups (castes and subcastes), which led to the creation of various horizontally organized groups that functioned at a variety of geographic and social scales. These groups had significant regulatory roles, affecting social behavior and economic

11

12


activities, as well as roles in regulating producers’ interactions with state and temple institutions. And, in certain circumstances, craft producers from individual or multiple hereditary communities were able to organize collective actions on a large scale, in ways that show both a level of political sophistication and a political awareness that is seldom acknowledged in studies of precolonial Asia.

2

Specialized craft production: archaeological approaches

The study of specialized craft production – its emergence, organization, and associated technologies – has a long and esteemed history in archaeological research (see Wailes 1996). Specialized production, of subsistence resources and durable craft goods, has been viewed as a defining characteristic of sociopolitical complexity and its study as key to understanding economic and sociopolitical structures of complex societies (particularly states and empires) and, increasingly, of so-called “egalitarian” societies as well (see Bayman 2002; Mills and Crown 1995; Sassaman 1998; Spielmann 1998, 2002). In complex societies, social, ideological, and occupational differences serve to distinguish among categories of people, and specialized labor becomes a normative means for the production and provisioning of certain categories of subsistence and other goods. States and empires in particular are characterized by a highly elaborated division of labor, based ultimately in the ability of communities or societies to produce subsistence surpluses that can be used to support individuals and communities who do not engage in food production. Factors underlying the emergence of specialized craft production have been explored by a number of scholars (e.g., Arnold 1976, 1985, 1993; Brown 1989; Childe 1951; Costin 1986; Rice 1981, 1991), and remain a pressing issue in the study of the nature and development of formalized systems of social inequality in human societies. Questions of ultimate origins are not, however, relevant to the study of Vijayanagara. Vijayanagara organization built upon at least 2,000 years of economic and sociopolitical complexity in South India. By the time the Vijayanagara state coalesced in the midfourteenth century AD, South India was characterized by a complex mosaic of hierarchically and horizontally arranged cultural groups, distinguished along geographic, linguistic, and religious dimensions and defined by kin, caste, and occupational identities. However, as discussed in chapter 4, these groups were far from fixed, and during the Vijayanagara period the constitution, status, and identities (both externally and internally defined) of diverse social groups – including craft producers – underwent significant changes.

14


Rather than focusing on origins, in this chapter I present an overview of recent research on the organization and technology of craft production and on the social, political, and ideological positions and identities of craft producers and their products. I do not intend this to be a comprehensive review of the voluminous craft production literature; nor do I seek to produce a cookbook of how to “do” craft specialization archaeologically (see Costin 2001 and in press, for recent review articles). Instead, I wish to highlight theoretical issues that are particularly relevant to the discussion of Vijayanagara craft specialization presented in this work. The actions of craft producers have, of course, important economic consequences. Craft producers manufacture goods that are consumed by individuals not directly involved in their production. Producers themselves often consume various raw materials and manufactured goods as part of the production process, and may rely on a variety of other specialists for goods and services. Thus, in the contexts of production and consumption of specialist craft products, individuals and groups are linked in complex webs of interdependence and interaction. The kinds of goods produced by craft specialists in any society and the scales of their production and consumption may vary widely – from valuables consumed by only very limited segments of the population to the more widely consumed goods of daily life. The spatial, technological, social, and political contexts of production and distribution may be similarly diverse, and within specific contexts, may vary between categories of goods and/or within particular technologies. Because much (though certainly not all) craft production leaves distinctive material traces, archaeologists are especially well positioned to address technologies, locations, and scales of production, and to examine their roles in the broader social, economic, and political contexts in which they occur. Along with considering issues of technology and economy in Vijayanagara craft production, I also view craft producers as social actors. That is, my focus is not only on the production and distribution of goods, but on the people who produced them and their broader roles and positions within South Indian society. Here I consider questions of recruitment, division of labor, and social identity, as well as the nature and range of craft producers’ relations with formal institutions, non craft-producing communities, and each other. As well as economic import, craft production and consumption have wide-ranging sociopolitical and ideological consequences and contexts. These can be approached from the perspectives of craft producers, craft consumers, and institutions. From the perspective of craft producers, we can explore the attribution of social status or social value to various producers

Specialized craft production

and/or acts of craft production. Does internal social differentiation exist among producers of particular kinds of materials? How do (or do?) producers interact in geographically extensive polities? How are producers of different categories of goods ranked relative to each other? Are these rankings shared or contested, and by whom? Under what conditions is social mobility possible for craft producers? Equally important to consider, and often affecting producers’ social and economic positions, is the sociopolitical significance(s) of craft products – the roles of (some) goods in defining and representing status, gender, class, religious affiliation, and other identities (including those of their producers). If we shift perspective – from craft producers to craft consumers and institutions – we come to two important issues. The first is the demand for craft goods. Demand affects the rates and contexts of consumption, and therefore, production. Consideration of demand obliges us to examine the functional attributes and contexts of use of craft goods, their exchange values and mechanisms for their distribution, as well as their social, political, and ideological significance (see Costin 2001). All of these may vary over time, space, and social context. A second topic relevant to addressing both consumption and institutional involvement in craft production is that of “control.” Control has, not surprisingly, been much discussed in studies of craft specialization, leading as it does to broader questions about the political and institutional structures through which it is exercised. Production can also be controlled or regulated by artisans responding to local conditions of resource availability and demands for their products. Despite or perhaps because of the frequency with which the term is used, control is a notoriously difficult concept to pin down, either archaeologically (see G. Stein 1998: 20–21; Costin 2001, in press) or theoretically. Control may be exercised in wide variety of ways with varying impact and consequences. Institutions may seek to limit access to raw materials or finished products; they may seek to regulate the form and quality of craft products, production processes and technologies; or they may control the bodies and lives of craft producers (see Clark and Parry 1990; Hayashida 1995: 13). Institutional representatives may directly and forcefully impose control and have the ability to implement a variety of coercive sanctions. Or, control over craft production may be more subtle and indirect, involving inducements and rewards rather than coercion. In the Vijayanagara case, the latter seems to be far more common than the former. Vijayanagara elites and institutions of various sorts were interested in craft production, but seldom in its direct management or coordination. Their control was exercised indirectly through taxation and, in the case of

15

16


state institutions, through being the best consumers of valued craft products such as elaborate textiles, and generous patrons of their producers. As a transformative process – of the natural to the cultural – crafting and its results often carry enormous ideological import that goes beyond the meanings attributed to the goods themselves. Both the act and the metaphor of crafting (see Helms 1993), of shaping and transforming the material world, often play powerful roles in assertions of sacred and political authority. Helms has highlighted this in her 1993 book Craft and the Kingly Ideal that explores the relation of specialized skills and exotic knowledge and claims to political and sacred authority. Among the examples she cites is the Mesopotamian Gilgamesh epic, which presents the Mesopotamian view of kings as the ultimate crafters, responsible for the production and reproduction of the social universe. Such views of rulers as crafters are widespread, also occurring, for example, among the Aztecs (Brumfiel 1998), the Visvakarma community of South India (see chapter 6), and the blacksmith-priests and other artisans of West Africa (Frank 1998: 5), among others. In my approach to Vijayanagara crafting, I take a very broad definition of craft and craft products, including some products and forms of production not commonly considered by archaeologists. I address ‘conventional’ crafts such as ceramics, textiles, metal goods, ornaments, and artifacts of other materials such as wood, feather, and stone, and paintings. I also include architecture and masonry as part of a continuum of stone working activities at Vijayanagara that ranged from quarrying and wall and reservoir construction to the carving of elaborately sculpted images on temple columns, palace basements, and bedrock boulders. Finally, I consider producers whose products were in words, sound, or movement – poets, bards, musicians, and dancers. I group all of these under the rubric of craft in that all entailed the investment of labor by skilled practitioners who labor to transform potential into finished products that were in turn consumed by non-producers. In the remainder of this chapter, however, I largely restrict my focus to the more standard archaeological domain, to consider how archaeologists have categorized, documented, and interpreted specialized craft production in complex societies.

T H E O RG A N I Z AT I O N O F S P E C I A L I Z E D C R A F T P RO D U C T I O N Several influential studies on specialized craft production published over the last three decades have defined the terms through which archaeologists have approached this topic. They include a number of important typologies


that focus on units of production – the size and structure of production groups – developed by van der Leeuw (1977, 1984), Peacock (1982), Costin (1991), and others (see Table 2.1 and following discussion). A second approach to categorizing craft production was presented by Brumfiel and Earle in 1987. They were particularly interested in exploring the relation between specialization and sociopolitical complexity, and approached this by highlighting factors that link political processes with productive organization. Brumfiel and Earle (1987) emphasized the distinction between the production of wealth goods and staple products (also D’Altroy and Earle 1985; D’Altroy 1992), and the different roles these play in political economies. Also prominent in their approach was an emphasis on how producers interact with political elites and institutions, that is, in the distinction between specialists who are attached to or independent of elites and/or institutions (Brumfiel and Earle 1987: 4), returning us to the question of control raised earlier. In their discussion, Brumfiel and Earle listed multiple parameters or dimensions of variation that are embedded in the concept of specialized production. These include the affiliation of specialists (i.e., independent or attached), the nature of the product, the intensity of specialization (i.e., partor full-time), the scale of the production unit, and the volume of output (1987: 5). Costin expanded on Brumfiel and Earle’s perspective in a highly influential article published in 1991, in which she outlined four parameters for documenting craft production: context, concentration, scale, and intensity. She has developed this perspective further in more recent works (Costin 2001, in press). I will turn to the dimensional perspective after a discussion of typological approaches.

Typologies of production units Specialized craft production can occur at a variety of scales and in a range of productive contexts. Van der Leeuw (1977, 1984) has presented a five-part classification of specialized production units (Table 2.1). Although focused on ceramics, his framework was intended to be generalizable to other spheres of craft production. Criteria for distinguishing among van der Leeuw’s five types include the size of the unit of production, the intended consumers, and the technology of production. The basal level in this hierarchical classification is household production, in which artisans produce for consumption within their own household, and in which all households in a community produce the same range of goods. While non-specialized in the sense in which the term is most often used in

17

18


archaeology, household production is minimally characterized by divisions of labor according to gender and age, such that not every household member engages in identical productive activities. Formal specialization appears in van der Leeuw’s category of household industry. Production still occurs in the household context, but the unit of consumption has expanded beyond the producer’s household to include other households in the community. Specialists typically have added craft production activities to their normal range of household tasks, although craft production does not provide their primary economic support. The third level of production is the workshop industry. Artisans in these contexts derive the bulk of their income from craft production, and produce at a larger scale and on a year-round (or nearly so) basis. Workshops may be spatially separated from households or may be located in household contexts. Often, nuclear or extended families comprise the basic unit of production, though some recruitment of non-household laborers may also occur. Some task specialization may occur among workshop members. The scale of production may be quite variable, as also the amount of effort devoted to craft production versus other economic activities by individual workshop members. Fourth is what van der Leeuw has termed the village industry. This is characterized by multiple producers and workshops located within a settlement, who produce to serve the needs of a larger region. Production may be on a relatively large scale, with concomitant development of task specialization, so that the various steps in the production processes are themselves specialized (van der Leeuw 1984: 756). Village producers may be associated with merchants or traders who distribute craft products beyond the village bounds, or may be directly involved in distribution themselves. Van der Leeuw’s village industry has some similarities to the category “village specialization” or “community specialization” that has also recently appeared in the literature (e.g., Habicht-Mauche 1995; Hegmon, Hurst, and Allison 1995; Welsch and Terrell 1998). Hegmon et al. (1995: 33) define community specialization as a context “in which individual specialists, aggregated in a limited number of communities, produce pottery [or other artifacts] for regional distribution.” In this case, while residents in certain villages may “specialize” in the production of specific wares, the scale of production may nonetheless still be relatively low, with production occurring in household contexts, as a household specialization or household industry. There are thus some marked differences from van der Leeuw’s category, which encompasses both the location of production and its scale.


Table 2.1 Typologies of the organization of craft production van der Leeuw

Peacock

Costin

Household production Household industry Workshop industry Village industry Factory industry

Household production Household industry Nucleated workshops Manufactory production Factory production Estate production Military and other official production

Individual specialization Dispersed workshop Community specialization Nucleated workshops Dispersed corvée Individual retainers Nucleated corvée Retainer workshop

Table compiled from van der Leeuw (1977, 1984), Peacock (1982), and Costin (1986, 1991).

The factory industry is the fifth and largest in scale of van der Leeuw’s productive types. Here, production occurs in workshops or factories that are physically isolated from household contexts, and involves large numbers of individuals. Often, these individuals are not related through kin ties, and may be recruited through a wide variety of means (wage labor, corvée labor, enslavement, etc.). In factory industries task specialization is high, as is investment in production facilities and raw materials. Emphasis on product standardization and quality control is also expected to be high, and such facilities typically include managerial personnel responsible for coordinating scheduling and evaluating production. These managers can include state or other administrative personnel (e.g., temple elites, military officials), merchants, or private entrepreneurs. Table 2.1 summarizes similar typologies of specialized craft production developed by Peacock (1982) and Costin (1986, 1991). Like van der Leeuw’s classification, their approaches are hierarchical and increase in scale from individual specialization at the household level to large-scale production in factory contexts, with concomitant increases in technology and time investment and output. Costin and Peacock’s classifications differ from van der Leeuw’s in their incorporation of information on managerial control to the consideration of the scale of the unit of the production in their definitions of types. While these additions are useful in thinking about who controls or directs craft production in particular contexts, it may be wisest to avoid mingling these different dimensions and instead adhere to the more restricted typology of van der Leeuw for considering units of production. These various typologies of craft production organization (see also Rice 1991; Sinopoli 1988) have proven useful in providing ways to think about questions of scale and the social units involved in specialist production.

19

20


As a frame for discussion they, like numerous other simplifying typologies archaeologists use, are valuable. Nonetheless, like other typologies, they can lead scholars to ignore or reject other possible ways of organizing human labor, or to reify categories that may have been in practice quite fluid (e.g., Feinman and Neitzel 1984; Rice 1991; Yoffee 1993). No doubt, numerous other approaches or definitions of units are possible. The frameworks presented in Table 2.1 are all hierarchical, proceeding from simple to increasingly more complex organizational forms (in numbers of participants, technologies, and webs of economic and political relations). There is an implicit evolutionary trajectory embedded in these models, in which scale of production units (and hence complexity of production) is positively correlated with sociopolitical complexity. This risky association posits relations that instead need to be evaluated in specific historical contexts. To what extent do we see a sequential development of various types and scales of production units, and under what circumstances? And, how are such developments associated with changes in larger political or economic structures? While household and individual production can be presumed to have temporal precedence over workshops and factories when considered at a global scale of historical change, the development of complexly organized forms of production is not inevitable, even in complex societies (see also Underhill 1991). Nor does the appearance of large-scale production units preclude the simultaneous existence of smaller-scale organizational forms. Thus, even when factories or other large-scale productive units come into existence, household or workshop industries may continue to produce a range of craft products. These may be entirely different than those produced in the larger facilities or contexts (e.g., factory-produced textiles and householdproduced ceramics), or may be similar, but produced for different consumers or in different contexts (e.g., urban vs. rural textile production). Examination of multiple craft products in multiple contexts is therefore critical. In addition, the development of certain forms of political organization (e.g., hierarchical societies or states) does not necessarily require parallel, linear changes in productive organization. That is, states do not equal or imply factory industries or other large-scale units of production, even in contexts of increasing economic differentiation and increased demands for craft goods. In fact, Feinman (1999) has argued that in early state periods in Oaxaca, many craft products were produced in household contexts, a pattern that continued in Mesoamerica into the Aztec period (Brumfiel 1998), and also characterized Vijayanagara.


Craft producers and institutions can respond in a variety of ways to heightened demands for craft products. These can include increasing the complexity and scale of productive units. However, they can also include artisan-initiated technological innovations that result in greater output per artisan or workshop, as well as increased specialization and greater interactions between small-scale workshops (for example, workshops that specialize in weaving interacting with workshops that dye yarns or woven fabrics). The linkages among workshops or producers can be facilitated by horizontal structures, such as guilds or other organizational forms, rather than by hierarchical, and institutional, ones. Artisans may also respond to increased demands by reducing labor investment, such as by decreasing the complexity or quantity of ornamentation, engaging in less labor intensive raw material processing, or by otherwise simplifying production techniques and resource use (such as a shift from decorated to undecorated ceramics, see Wengrow 2001). We might also see an increase in the overall number of production units rather than in their size or organization. In sum, in most societies, and especially in complex societies, it is reasonable to expect the coexistence of multiple forms or relations of productive organization. And it is the differences among these various productive modes and the reasons for them that raise far more interesting questions than the necessary first step of cataloguing the productive organization for individual categories of craft goods. The Vijayanagara case provides a good example of the dangers of making simple assumptions about the organization of craft production in complex political systems. By any measure, the Vijayanagara empire was complex. Its geographic extent was vast, and the empire’s population in the sixteenth century has been estimated at around twenty-five million (B. Stein 1989a, see chapter 4). South Indian society was highly differentiated along ethnic, caste, religious, political, and economic lines. A wide range of craft goods was produced and consumed, ranging from luxurious textiles and gold ornaments to earthenware ceramics and wood and iron agricultural tools, and there is good evidence that demands for textiles, metal objects, and many other craft goods increased dramatically throughout the Vijayanagara period. Virtually all of these goods were made by specialists, who differed in their economic, social, and political status and mobility. However, with some exceptions, craft production occurred within household contexts, in workshops whose membership was primarily defined by relations of kinship and marriage. Craft producers were linked by caste membership into collectivities of varying geographic extent, that could, in some cases, act as corporate units; producers also formed larger inter-caste affiliations, which also served

21

22


regulatory roles and cooperated in acts such as social protests (see chapter 7). Neither factory industries and other large-scale units of production, nor hierarchical mechanisms of control, were instituted in response to increased demands for craft products in the Vijayanagara context (see also Sinopoli 1998, 1999, and chapter 6). It may be argued that Vijayanagara was aberrant, an exception to more normative craft production strategies that characterize most state-level societies. Certainly, we do see the appearance of large-scale production structures in many other early states and empires. However, I suspect that the South Asian response is more common than has been recognized; like Feinman (1999), I suggest that scholars need to more carefully explore the broad range of craft production regimes that can exist in states. Doing so forces us to reconsider long-held assumptions about craft production and labor structures in state societies, as well as our assumptions about states, particularly concerning the extent of their managerial authority over diverse economic (and other) realms, and the presumed effectiveness of state systems in breaking down kin-based economies, such as household workshops. I will return to these questions in chapter 8, after I present the Vijayanagara data.

Parameters of craft production In an effort to redirect considerations of craft production from typological to organizational concerns, Cathy Costin (1991) and Christopher Pool (1992), following Brumfiel and Earle (1987), separately suggested that scholars focus their attention on multiple parameters that play a role in productive organization. In 1991, Costin highlighted four such parameters: scale, concentration, intensity, and context. Pool defined somewhat different parameters for study – scale, intensity, efficiency, size of production unit, segregation of activities, location of production, and variability of products. All of these parameters are implicit in the classifications presented above and are found in many discussions of craft specialization. However, by focusing on the variability in each, the authors hoped that we could explore a wider range of productive arrangements than permitted by the typologies, as well as consider the various factors that underlie them. Costin’s 1991 publication has been arguably the most influential and widely cited general discussion of craft specialization of recent years, and has provided a framework that has been acknowledged, if not fully applied, in numerous studies. Here, I briefly summarize her initial discussion, and then


consider how additional theoretical concerns have influenced and expanded approaches to the study of craft production since 1991. As noted, Costin’s initial formulation provided four parameters for the study of craft production: scale, concentration, intensity, and context. Costin’s focus was on the social, spatial, and political settings and structures of production, rather than the goods produced or their uses. She viewed each of the parameters as continua, which could vary independently of each other. Thus, rather than a limited set of types, craft production could be structured in a multitude of ways, within a single culture and cross-culturally. The parameter scale most closely corresponds to the units of production discussed earlier, and refers to both the sizes of production units and the rules or principles for their recruitment. Concentration refers to the spatial distribution of specialists across the physical landscape and in relation to other producers and potential consumers. Intensity is defined as the “amount of time producers spend on their craft” (Costin 1991: 16), with full-time specialization at one extreme of the continuum of variability and part-time specialization at the other. In the latter, craft production is only one of a set of productive activities engaged in by specialists, whereas for full-time specialists, production of craft goods is the primary or exclusive economic task engaged in. The fourth parameter, which Costin termed context, refers to the relationships of productive activities and producers to sociopolitical institutions and/or patrons. At one end of this continuum is what Costin, following Brumfiel and Earle (1987) and ultimately V. Gordon Childe (1936, 1951; Wailes 1996: 3), referred to as attached specialization, in which institutions or elites directly regulate some aspects of craft production. Such regulation may, as noted above, include control over raw materials or tools of production, control over access to and/or quality of finished products, or control over producers themselves, who may labor as part of elite households or in institutionally sponsored and supervised workshops. At the other end of this continuum is independent specialization. Here, production is not directly controlled by institutional authorities or elites, but is regulated by autonomous artisans or groups of artisans who work largely independently of political institutions, and whose production activities are influenced by consumer demands and competing economic priorities. Independent producers may be involved in political economies (i.e., they may be required to pay taxes or tribute related to their production activities), but they have greater flexibility in the scheduling and organization of their productive activities and the disposition of their products than do attached specialists. Costin’s concept of context, and of attached and independent specialists, have generated more discussion than

23

24


any other topic she discussed, and I will return to this issue in more detail below. Costin’s parametric approach to the study of craft production provided a valuable framework for disarticulating the different activities and spatial and social contexts of craft production activities. Each of the parameters she presented poses challenges to archaeological research, and not all aspects are easily accessible from archaeological evidence, even in those cases where we recover production locales. For example, scale, or the size of production units, may often be under-represented both archaeologically and in the ethnographic literature by an emphasis on only a limited set of the activities involved in crafting particular products. Thus, Rita Wright (1991) has pointed out that discussions of pottery production in South Asia have emphasized the role of males as potters, while ignoring or underplaying the contributions of other household members. In particular, the roles of women and children in raw material collection, hand building, decoration, and firing are ignored, because pot making is defined in terms of the use of the potter’s wheel, an exclusively male activity. Similarly, when addressing concentration, or the spatial distribution of specialists, it is important to recognize that many activities associated with craft production take place outside of workshop contexts (for example, raw material or fuel acquisition). In addition, for some craft activities, different stages of the production process may occur in different locales. Kenoyer, Vidale, and Bhan (1991) have documented such a pattern in their ethnoarchaeological study of carnelian bead manufacture in Khambat, Gujarat, in western India. Here, subsidiary workshops carried out various stages of the bead-making process under the supervision of large central workshops from which raw materials were distributed and to which finished products (and waste) were returned. The amount of time invested in craft production or intensity is also extremely difficult to estimate with any degree of precision, as definitions of appropriate labor allocations and of full-time or part-time labor are culturally specific (Rice 1991: 263–264).

F U RT H E R C O N C E R N S I N T H E S T U DY O F C R A F T P RO D U C T I O N As noted earlier, Costin’s initial framework focused solely on craft production. She did not explicitly consider the goods produced by specialists nor their consumption. Recently, Costin (2001, in press) has expanded upon her original four parameters to argue for a fuller examination of “craft production systems.” Such an examination seeks to both describe and explain


the nature and interrelations of six critical components: artisans, the means of production (materials, tools, knowledge, and skill), the organization and social relations of production, the objects produced, relations of distribution, and the consumers of craft goods. In this, her work builds on salutary recent developments in the study of technologies, consumption, political economies, and material culture, as well as new interests in the roles and status of non-elite communities, such as (often) craft producers in state societies. Below, I address several of these themes that are relevant to my discussions of Vijayanagara. These include the study of artisans as social actors and of craft production as a social process, new theoretical approaches to material culture, and issues related to political economy, attached and independent specialization and mechanisms for wealth mobilization and distribution in state societies. I will not discuss archaeological methodologies for the study of craft specialization in any detail. A broad methodological overview is presented in Costin (in press), addressing characterization studies (also Neff 1992; Tite 1999); measures of product standardization (see Arnold and Nieves 1992; Arnold 2000; P. Arnold 1991; Benco 1988; Blackman, Stein, and Vandiver 1993; Longacre 1999; Rice 1991; Sinopoli 1988; Stein and Blackman 1993); measures of labor investment or production steps (e.g., Feinman, Upham, and Lightfoot 1981; Costin and Hagstrum 1995); and other relevant archaeological techniques. I will address some of these as appropriate in my consideration of specific Vijayanagara crafts.

Artisans as social actors The production of craft goods requires resources, (more or less) skilled practitioners, and knowledge. Until relatively recently, most studies of craft production have focused on general features of technology (e.g., the presence of wheel-made vs. handmade ceramics) and on the organizational structure of productive units (particularly, their relations with state institutions). Changes in these dimensions were viewed primarily as technological and organizational responses to external stimuli, whether the demands of political elites or of market processes and consumers or changes in raw material access or availability. Less attention has been paid to the producers themselves, as active participants in and shapers of productive systems and relations as well as of the goods they produce (see also Costin 1998a: 4–5, 2001). In this section, I briefly address some of the social dimensions of craft production, looking both at production as a social act and at craft producers as participants in the broader social and political contexts in which they ply their trades. Throughout, my emphasis is on variability and diversity. Such

25

26


variability may be particularly true of a highly complex, highly elaborated political economy such as Vijayanagara’s, but it is also likely the case, albeit at a smaller scale, in far less complex societies. As the discussion of individual crafts in chapter 6 illustrates, it is not appropriate in the Vijayanagara context to think of producers, even of a single type of product, as unitary. Weavers who resided in villages and made garments for their neighbors lived very different lives and had very different opportunities than weavers who resided in temple towns or urban centers and made cloth for imperial nobility and other elites. South India’s weaving communities differed also in technology, materials used (e.g., silk, cotton, wool), and social affiliation, including language, religion, caste, and ethnicity. Some Vijayanagara craft producers may have moved between different productive systems or relations across the course of their lifetime or over a single year. For example, mobile blacksmiths who forged hoes and ploughshares for agriculturalists may have worked at other times as laborers in imperial military camps manufacturing or repairing armor and weapons. In these cases, the technology employed was more or less constant, but the physical settings and institutional relations in which producers labored differed dramatically. For most archaeologists, working without the benefit of written records, many of these varied and co-occurring relations can be difficult or impossible to identify. In the Vijayanagara context, historical sources provide us with a richer understanding of the complexity and diversity of producing communities, and in some cases even of specific individuals, than available to prehistorians. In many discussions, craft production has been viewed as responding to external stimuli – i.e., wheel-made pottery is a response to demands for increased output through increased productive efficiency. This is indeed a useful perspective – producers do react to the external conditions of their world and modify their technologies and production practices accordingly. Yet in such models, the producers themselves are strangely silent: the production system reacts, but the producers involved are merely pulled along in larger systemic flows. They have not been viewed as active, thinking, individuals and groups who make the decisions that affect the nature and form of these reactions. By refocusing our perspective on craft producers, we still necessarily view them in relation to institutions and external conditions, but we also see them as individuals and communities whose decisions and responses help to determine their social and economic positions, as well their products. The embedding of craft production in the social can be explored at multiple levels – from a focus on the production process itself as a social act, to a consideration of social relations and social values within


units of production, between various categories of producers, and between producers and their broader social and cultural milieu.

The social foundations of technological practices Acts of craft production, even in those relatively rare cases when a lone artisan carries out an entire production process, are ultimately social. Knowledge concerning how goods are produced and the appropriate forms that they should take is transmitted through interaction and learning – between parent and child, master and apprentice, employer and employee. Understanding how artisans are recruited, how knowledge is transmitted, and how it becomes internalized through practice has received considerable attention in the archaeological and ethnoarchaeological literature over the last few decades. The internalization of production techniques has been explored using the chaˆıne-opératoire approach developed by Leroi-Gourhan (1964). This largely descriptive technique seeks to document the complete sequence of steps and physical gestures entailed in acts of production. The execution of some of the production steps may involve conscious decisions by artisans; others may involve rote behaviors that are consciously learned as an artisan masters her or his craft, but later become highly routinized and largely unconscious. Recently, a number of archaeologists have sought to address the social context of technological knowledge by merging the precision of the chaˆıne-opératoire with French sociologist’s Pierre Bourdieu’s (1977) concept of habitus, also concerned with the dialectic between cognized and non-cognized aspects of cultural behavior (e.g., Dietler and Hiebricht 1998; Dobres 2000; Dobres and Hoffman 1994, 1999; Gosselain 1998, 2000; Lemonnier 1989, 1992; Roux 1989; Roux and Marasso 1999; Schlanger 1994; van der Leeuw 1994). Other scholars have combined an interest with technological sequences with anthropological approaches to style and “practice theory” into an approach that has come to be referred to as “technological style” (see Hegmon 1998; Lechtman 1977, 1993; Stark 1998). Both of these approaches address both “ways of doing” (i.e., technologies of production) and the social and cultural contexts in which they are mastered and reproduced.

The social contexts of production If transmission of technological knowledge is a social process, acts of specialist production also require social groups of various sizes. These include the diverse participants in the production of any particular craft

27

28


product – from raw material procurers or producers (who may include other specialists, e.g., miners, charcoal producers, and smelters who provide ores and smelted metals to forgers or casters), to people involved in various stages or aspects of the production process, to those involved in product distribution. Producers also interact with those who acquire and use their products, whether elite patrons or market customers. The nature and kinds of social and economic relations that develop among producers and between producers and consumers can vary widely, and for a wide variety of reasons. And both the natures of these relations and the reasons for their patterning are important questions for study that can allow us to consider both specific historical cases and broader cross-cultural patterns. Many of these questions have already been alluded to and will be explored in greater detail in discussions of Vijayanagara crafts and crafters in subsequent chapters. Here, I simply provide a brief outline of some issues that I will return to later. Considering first the producers themselves, questions of interest can include those related to the organization of production, discussed above, including the size and structure of craft-producing units, the division of labor, and how labor is recruited. Production units may be formed through kin relations, marriage, and heredity, or laborers may work for wages, or be “drafted” or enslaved by institutions or elites. Gender and age, as well as skills and knowledge, can affect or determine labor activities and the values accrued to specific occupations or tasks. We can also explore whether or not there is potential for social mobility within and/or between occupations, and how this varies over time and over space (e.g., between rural and urban settings; see Brumfiel 1998). In addition, we can consider whether and under what circumstances communities of producers can or do act as effective corporate units. The presence or absence of guilds or other kinds of corporate organizational structures, the ability of communities of producers to negotiate for increased economic compensation or other privileges, and their ability to resist efforts at increased extraction of goods or revenues from producers, can all be addressed.

Status, meaning, and beliefs Social meaning or status may be ascribed to acts or knowledge of craft production, to communities of craft producers, and to craft products in a variety of ways. For example, both goods and producers may be highly valued, with high social status, as Weiner (1976) has suggested for the production of high-status textiles by elite Trobriand women (also Inomata 2001 on Maya


elite production). In such contexts, the qualities of the producers (as elites or possessors of special powers or knowledge) are also transferred to their products. In other contexts, craft goods may be highly valued while their producers are not, such as the enslaved and disenfranchised women who produced textiles in the palaces and temples in Ur III Mesopotamia (Wright 1998; Waetzoldt 1987). Value and status attributed to producers can be difficult to detect archaeologically in the absence of written documents. Mortuary data, iconography, and information from producers’ habitations or production facilities may provide evidence to examine such questions. Being a craft producer is, in many societies, an essential dimension of individuals’ self-perceptions and sense of social position. There are many historically and ethnographically documented examples in which the knowledge and skills of crafting are linked with particular sacred or ritual qualities. Lass (1998: 26), for example, has noted that in precontact Hawai’i, producers of highly valued crafts believed that their skills came directly from the gods and ancestors, and that their knowledge provided them a unique social and sacred status. South Indian Visvakarma metal workers (see chapter 6) viewed their craft as allied with the divine “architect of the world,” and placed themselves above Brahmans in social and ritual status (there is however, no evidence that Brahmans or other communities agreed with this assessment). Along with sacred associations, crafting may be linked to other basic characteristics of social identity, such as gender. Thus, throughout prehispanic Mesoamerica and South America, the production of textiles was viewed as an essential attribute of femaleness: to be a woman was to weave and, in Mesoamerica, to spin. Among the Aztec, for example, infant girls were given spindle whorls and weaving battens (Brumfiel 1991; McCafferty and McCafferty 1999), and female deities were often depicted with spindle whorls. It is, however, dangerous to assume a priori that gendered or other socially restricted occupations will necessarily be marked or ideologically charged. While some crafts or activities may become closely linked with various dimensions of social or individual identity, others may not be, however widespread they are in practice. Costin (1998b: 133) has noted that while weaving was a key component of female identity among the Inka, and several named categories of weavers are known, no term for spinners or spinning appears in early Quechua dictionaries. Evidence suggests that spinning was ubiquitous, and was practiced by all women as well as by some men. However, this particular craft was not marked, at least linguistically, and Costin (1998b: 133) has suggested that this activity was “largely invisible in the scheme of social recognition and importance.” It was the manufacture of

29

30


textiles from spun thread that was deemed essential to cultural conceptions of identity, while spinning, the necessary precursor to weaving, was not. In the above examples, crafting is understood in terms of broad social categories such as gender. In certain contexts and for certain crafts, the individuality or uniqueness of the object may be of key importance. This is particularly the case for goods with high social or ideological status. In some contexts, the unique skills or qualities of individual producers may also become worthy of notice. Where it is possible to distinguish the products of individual artisans or workshops, we can explore questions concerning both the social positions of those producers and the relative values ascribed to their work. That is, in what contexts are the contributions of individual “masters” considered worthy of record; and how do these differ within and between societies? This is not solely a question of skill or distinctions between the western concepts of “art” versus “craft,” though certainly skill and the cultural and economic value of craft products play a role. It may also be tied into the ideological values attributed to craft production. For example, Wright (1998: 60) has argued that in Mesopotamia, crafting was considered a defining attribute of both kings and gods. The producers of goods, even those most valued, were non-nobles and could be seen only as fabricators rather than crafters, and were not named or recognized as the creators of specific products. In the Vijayanagara period also, it is relatively rare to learn the names of individual sculptors, painters, or architects, despite the high quality of some of their products. In contrast, in other traditions, such as in parts of Europe at the same time, the name of the particular artist was often a key component of the value of an object, and signatures were affixed to many categories of highly valued goods. Any consideration of craft producers as social actors requires that we explore a number of relevant factors. We can expect a complex, dialectical relation between the social and economic status of craft producers, the ideological values attributed to producers and production, and the abilities of producers to affect or determine the conditions of their own existence. These dimensions are, of course, also bound up in broader societal factors and conditions – political organization, social stratification, and religious beliefs, among others. Finally, the instrumental function, value, and meanings of craft products in particular cultural contexts must also be considered.

Material culture It is perhaps surprising to highlight material culture studies as a new concern in archaeology and craft production studies, since archaeology is by


definition the study of the material. In that sense, studies of material culture have indeed always been integral to archaeological analysis, particularly for interpreting chronology and, to a lesser extent, artifact function and use. However, these relatively narrow perspectives have expanded greatly over the last few decades. In the late 1960s and early 1970s, studies by processual archaeologists, such as Deetz (1965), Hill (1970), Longacre (1970), and others, began to analyze material culture (specifically, ceramics) to examine the spatial distribution of social groups. Although na¨ıve in their assumptions about a straightforward (and largely unconscious) association between ceramic decoration and kin groups, these studies opened up a new horizon for studying the social through analyses of archaeological remains. Since this early work, there has been a tremendous growth of interest in theoretical and methodological approaches to material culture within archaeology, anthropology, and social theory. While maintaining a focus on the economic and instrumental properties of artifacts and landscapes, these approaches also focus on the symbolic and phenomenological dimensions of goods – in defining and mediating social interactions, symbolizing and shaping various social identities, and reproducing, representing, and resisting social and economic inequalities (see, for example, Appadurai, ed. 1986; Chilton, ed. 1999; Glassie 1999; Robb, ed. 1999; Schiffer and Miller 1999; Stark, ed. 1998; Thomas 1991). The centrality of material culture in virtually all human activities and interactions is perhaps so obvious as to be a truism. However, recent studies have somewhat shifted focus from looking at goods as being shaped by humans to looking at the role of the constructed world in dialectically shaping human perceptions, knowledge, and behavior. Incorporating understandings of the social role of goods in contexts where goods are produced by specialists for use by others requires appreciation both of the complex relations between production and consumption and among various systems of meaning and knowledge in particular cultural contexts. All goods have meanings, but any good may simultaneously have many different kinds of meanings to many different individuals (and even to a single person). Producers have both technological knowledge and knowledge of the symbolic meanings of production acts and processes, and of finished goods. The knowledge and perceptions that consumers have of the same goods are likely to be quite different, especially as they are further removed from the context of production (perhaps most dramatically, in situations of culture contact: e.g., Thomas 1991; Saunders 1999). Further, neither knowledge about nor the meanings of goods are constant. They change over time and over space, and across the life histories of individual objects (for the latter see Kopytoff 1986; Moreland 1999; Lillios 1999).

31

32


C R A F T P RO D U C T I O N A N D P O L I T I C A L E C O N O M Y I N S TAT E S A N D E M P I R E S A major focus of my work concerns the relations between craft production and political and institutional structures. As noted earlier, archaeologists and social theorists have long argued for an inextricable relationship between economic differentiation and political complexity. Craft products assume a variety of roles in political economies, as prestige goods or symbols of power and status, exchange goods, sources of wealth, and as tools for activities that produce the material infrastructure of states and empires, including subsistence production, mining, transport, and military endeavors. Craft producers and institutions can exist in manifold relations, with varying degrees of autonomy and interdependence, and much production can occur outside of the view or control of the state. In this section, I address two major issues relevant to the study of the political economy of states and empires: the relations of artisans to institutions, and the roles of craft goods in political economies.

Specialists and the state: attached vs. independent specialization The distinction between attached and independent specialization drawn by V. Gordon Childe (1936, 1951), and later Brumfiel and Earle (1987) and Costin (1991) in her parameter of context, has been taken up by many scholars seeking to explore the political dimensions of craft production activities (for example, see contributions in Wailes, ed. 1996; Brumfiel and Earle, eds. 1987; Costin and Wright, eds. 1998; also Peregrine 1991). However, adoption of this concept has been neither easy nor straightforward, and the characterization of attached and independent specialization, along with its archaeological identification, have proven to be very complex (G. Stein 1996, 1998). In her 1991 formulation, Costin argued that context, like her other parameters, should be understood as a continuum of variability. There is a dualism here, in that attached specialists are viewed as being dependent upon and tethered to institutions, while independent specialists exercise greater choice over their activities and labor. However, considerable variability can lie within each of these relations. Thus, attached specialization can include: (1) intensified household production, in which kin-based units produce in excess of their needs to satisfy institutional demands; (2) dispersed corvée production, in which a portion of producers’ production


activities in local settings is reallocated to the state; (3) individual retainer production, where individual skilled artisans are recruited to produce for elite patrons or state institutions; (4) nucleated corvée, in which part-time labor is recruited by the state to work in administered settings or facilities; and (5) retainer workshops, large-scale spatially segregated workshops with full-time artisans (Costin 1996). While Costin envisioned a wide variety of relations between sociopolitical institutions and producers, others have been more rigid in their application or conception of the attached vs. independent distinction. As a result, the concept has become both highly reified and, more recently, something of a “straw man,” subject to considerable criticism. In particular, the distinction has been criticized for not adequately addressing the wide variety of relations that can exist between producers and patrons (e.g., Lass 1998; Childs 1998; Inomata 2001). Such relations include, as Ames (1995), Helms (1993), Inomata (2001), Marcus (1987), Reents-Budet (1994, 1998), and Weiner (1976) have reminded us, the fact that members of elite households and status groups may themselves be craft producers, typically of highly valued and restricted prestige goods. That there is a wide variety of relations between producers and institutions is neither surprising nor unexpected (and indeed is explicit in Costin’s formulation); however, the repetition of this critique in the literature suggests that there is something problematic in how archaeologists have conceived of or applied these concepts. I think that the problem lies in the tendency to perceive attached and independent specialization as states of being or fixed and exclusive relations and identities, rather than as particular relations of production with specific, and often limited, duration (see also G. Stein 1996: 25–26; Kenoyer, Vidale, Bhan 1991: 46). There are certainly examples where relations of attached specialization produced structures of long duration that involved the vast majority of producers’ labor over long periods. These belong primarily to Costin’s “retainer workshop” and “nucleated corvée” production units and include, for example, textile and other craft workshops of Ur III Mesopotamia known to us from texts (a short-lived period of unusually strong state centralization; see Wright 1998; Steinkeller 1987; Waetzoldt 1987). Ur III state-controlled workshops employed dependent and coerced enslaved, corvée, and “semifree” laborers to produced textiles and other valued goods for the palace or temples. Laborers in such facilities were not free to leave the workshops, nor did they retain any control over the products of their labor. Instead, the cloth they produced belonged to the institutions for which they worked, and the producers were dependent on sponsoring institutions for rations of foodstuffs, clothing, and in some cases, land.

33

34


Inka textile production also involved attached specialization (Costin 1998b) by aqllakuna, high-status women who resided in segregated quarters, and by qompikamayoc, male weavers who resided in state-sponsored colonies (formed through forcible resettlement; Murra 1962, 1980, 1989; Rowe 1979). Here, too, the coercive dimensions of the attached relationship and dependency of the producers were of major significance, but the ideological underpinnings, especially for the aqllakuna, were dramatically different. Their title “chosen female” referred to women of high social status and rank who were viewed as part of the royal household, metaphorically defined “as wives of the state” (Costin 1998b: 134–135). Aqllakuna who engaged in weaving produced high quality textiles used in state rituals and redistribution. The formal and durable institutional framework of production in both the Ur III case of low-status laborers and this Inka example of high-status craft producers, may provide evidence for context where being an attached specialist does seem to have been a primary component of a producer’s identity, for at least some stage of their lives. While these examples of attached specialization are important, it is, however, critical to recognize that these kinds of attached relations seem in fact to be relatively rare historically, and they should not constitute our defining model of what is meant by attached specialization. Other kinds of less enduring or defining attached relations may in fact be more common in many societies and contexts. It is also the case that the kinds of attached productive relationships discussed above, where dependent artisans labor in spatially segregated and materially distinctive production contexts, are theoretically among the more straightforward for archaeologists to identify on the basis of variables of location, layout, and artifact content of production facilities. It is therefore not surprising that this variant of attached specialization has received the greatest emphasis in the literature (and has been found wanting). However, it is important to recognize that even in formal workshop contexts, the kinds of relations that may have existed between producers and institutions can vary widely, and can change over time. For example, artisans at the Chinese imperial porcelain kilns at Jingdezhen produced ceramics for the imperial court for much of the year, and during those periods potters were supervised by and “attached to” imperial officials. When the official demands were filled, the Jingdezhen potters produced similar wares for market distribution, using the same kilns, technologies, and personnel (Hobson 1962; Hayashida 1995: 17). The material signatures of kilns and production debris would be identical for the periods of independent and attached production, as would many (although not all) of the ceramic products manufactured (though


they would differ in the presence or absence of an imperial mark on the vessel base). But the political and economic contexts of ceramic production were radically different. Other variants of attached production relations can be even more difficult to distinguish based on archaeological evidence alone. In many contexts, as in Jingdezhen, specialists may shift in and out of attached or independent production relationships, sometimes in the same production settings and employing the same personnel and technologies. Other examples might include textile production by Aztec women, who devoted a portion of their efforts to meeting obligatory tribute demands, while also producing goods to meet the needs of their household and to distribute through market or other mechanisms. Whoever they produced for, production occurred in household settings using a consistent technology; only the intended consumer varied. The court painters of Renaissance Italy, like the poets of Vijayanagara discussed in chapter 6, illustrate another kind of relationship of attached production (for at least some of their products, see Gilbert 1998). The skills and reputations of individual artists were often in high demand by royal courts, religious elites, and wealthy merchants. For these patrons, the ownership of works of art served to enhance their “honor and splendor,” and the relations between patrons and artists were negotiated rather than coerced, often at terms quite favorable to artists (Gilbert 1998: 446). A similar situation may have existed for elite Maya vase painters (Reents-Budet 1994, 1998: 73–74). In contexts where the uniqueness and quality (however defined) of craft products were highly valued, and where skills and artistry of individual producers or workshops can be distinguished, artisans or groups of artisans may exert considerable sway in the terms of their relations with elite patrons. In sum, despite, or perhaps because of, the variability encompassed in the concept of attached specialization, it and its counterpart “independent specialization” remain valuable parameters to consider when examining relations of craft producers to institutions. However, it is also clear that stopping the analysis at an assertion of attached or independent specialization is not sufficient to describe the precise nature of the relation that existed between specific producers and patrons.

Craft goods in political economies Craft goods play diverse and critical roles in the economies of states and empires: as commodities for exchange, tools for subsistence or other

35

36


production, or as prestige symbols and sources of wealth. Studies of prestige goods in political economies have emphasized their role in inter-elite alliance formation, gifting, and wealth accumulation and creation of relations of social obligation (e.g., Friedman and Rowlands 1977; Junker 1999: Bayman 2002). Discussions of “prestige good economies” have largely considered pre-state societies, which are believed to lack the institutional means to penetrate as deeply into the full range economic practices and resources as can states and empires. This is not to say that prestige goods are not important in states as well. In fact, Terence D’Altroy and Timothy Earle (1985; D’Altroy 1992) have argued that a useful heuristic division can be made between state economies that emphasize the production, regulation, and access to wealth goods (“wealth finance”) and those that emphasize the production and distribution of staple products (“staple finance”). Theirs is, of course, a very general model, and most polities practice some combination of wealth and staple finance. In addition, definitions of wealth or staple goods are culturally specific and the lines between them may be difficult to draw. Nonetheless, differences among polities (and over time within polities) in the emphases placed on various kinds of commodities, and in the kinds of goods that are the foci for state control, can provide valuable insights into the economic foundations of particular states and variation among them. A further factor to consider in evaluating the role of craft goods in state economies is the extent to which products can move between various economic spheres, through market exchange (as among the Aztecs; Berdan 1994; Brumfiel 1980; Hassig 1985; Hodge and Smith 1994; Zeitlin 1991) or monetary transfers. Inscriptional records from the Vijayanagara period provide evidence for the expanded use of currency – gold, silver, and bronze coinage – in state administration and economic transactions. Taxes were increasingly collected in cash instead of goods; currency played an increasing role in temple donations, payment to military and other laborers; and at least some raw materials and finished craft products were exchanged through market mechanisms. A monetized economy coupled with formal systems of record keeping also allows the transfer of raw materials or craft goods to be displaced over time or distance. The Vijayanagara political economy was not fully monetized. Some taxes continued to be paid in kind; laborers associated with temples or institutions could be renumerated in foodstuffs, land, and other goods as well as or instead of cash. Some artisan workshops also maintained obligatory exchange relations with other households in their residential community; the ayagar or village servants (see pp. 100–102) included artisans who provided


craft goods to households and temples within their settlements, in exchange for subsistence and other goods. However, the fact that monetary transfers were possible opened up a range of possibilities for state finance and resource movement, not possible in economies that lacked such potential, such as those addressed by D’Altroy and Earle’s wealth and staple finance framework. In this work, my emphasis is as much on craft producers as on the specifics of their products. Where possible I will incorporate information on both the physical attributes of craft goods and their cultural meanings. Before I turn to an exploration of Vijayanagara craft and craft products, some additional background is necessary. In chapter 3, I contextualize this study historiographically, through an exploration of perspectives on the Asian state, and a consideration of how understandings of the state intersect with approaches to craft production in South Asia and elsewhere. As I noted in chapter 1, it is this through this lens that both the historiography and contemporary understandings of Vijayanagara polity and economy must be understood. In chapters 4 and 5, I provide background on the history of the Vijayanagara empire and the diverse sources of historical and archaeological knowledge used in this study, before turning to the topic of craft productions in the remainder of this book.

37

3

The South Asian state

From the fourth century BC until today, the “traditional Asian state” has occupied a distinctive place in the western imagination. While the location of the Asian other has shifted – from Persia to China and, during the early modern period of European colonial expansion, to India – the idea of Asian polities as distinctive, opposed, and less than, the states of Europe has remained a powerful and pervasive trope for more than two millennia. Through a series of transformations, the idea of oriental tyrants common in classical Greek historical and political thought became incorporated into the political thought of Renaissance Europe, Enlightenment philosophers, and eighteenth- and nineteenth-century British political economists, as well as of Marx, Engels, and their intellectual descendents’ models of the “Asiatic Mode of Production.” In the seventeenth through nineteenth centuries, understandings of Asian states and their perceived contrasts with those of Europe became a means through which diverse European social thinkers and philosophers, as well as Asian scholars and political activists, sought to better understand and transform their world. Understandings of the Asian state as despotic with pervasive centrally-controlled economies also made their way into the emergent discipline of archaeology, particularly in the construction of models for the Mesopotamian state (e.g., in the writings of Childe and Wittfogel, see pp. 47–49). These models in turn influenced subsequent archaeological approaches to other early states and empires, including ideas about craft production. Within South Asian historiography, most scholars no longer support models advocating “Oriental Despotism” or “the Asiatic Mode of Production.” Nonetheless, the distinctiveness of the Asian state continues to be emphasized in ways that reflect the legacy of these concepts (see also Inden 1990; Perlin 1985). Vijayanagara in particular and South India more broadly have not been central to longstanding discussions of the South Asian state. These have largely been based on interpretations of North Indian imperial polities and their successor states, especially the Mauryan and Mughal empires (Perlin 1985). However, the 1980 publication of Peasant State and Society in Medieval South India by historian Burton Stein and his


presentation of the segmentary state concept (discussed below) dramatically brought tenth- through sixteenth-century South India into the forefront of renewed debates about state organization. I have several goals in this chapter. The first is to trace the intellectual trajectory of the concept of an “Asian state,” with a particular eye to how the view of such a category of polity as other (and explicitly or by implication, less than) has conditioned South Asianists’ understandings of the region’s past. I see this as more than merely a historiographic exercise of relevance to only a small community of Asianists (many of whom have, after all, resolved these issues in their own minds). Instead, I lay the groundwork for an exploration of the impact of these long-lived concepts on interpretations of Vijayanagara history and political economy, and of precolonial India more generally. I begin my overview of writings on the Asian state with the ancient Greeks and conclude in the present, with the writings of several scholars who have sought to challenge these dominant perspectives, particularly through examination of later South Asian imperial states including Vijayanagara. Throughout, I pay particular attention to interpretations of political economy and, where information is available, craft production. Second, I consider how views of the distinctiveness of the South Asian state have affected the region’s position in comparative studies of pre-modern states and empires. For nineteenth-century scholars, including Hegel and Marx, historic South Asian states could only be viewed as stagnant relics of the world’s earliest civilizations (see Perlin 1985: 111; also Talbot 2001); for many later South Asianists, the uniqueness of the region, and particularly phenomena such as the “caste system,” placed South Asia beyond the scope of any comparative frame whatsoever. Both of these perspectives are problematic. The first denies the possibility of dynamic historical change from within South Asia. As a result, migration or foreign invaders must be invoked to explain any changes that did occur (see also Chakrabarti 1997; Morrison 1994: 190). The second perspective closes off scholarship on the region, and particularly its later precolonial history, from participation in broader theoretical debates within anthropology or history. In contrast to both perspectives, I make an argument for the comparative method and, specifically, for situating our examination of Vijayanagara both in the historical particularities of fourteenth- through seventeenth-century southern India, and in theoretical understandings of imperial states more broadly (Sinopoli 1994a, 2001a; Alcock et al. 2001). Third, and perhaps paradoxically, while South Asia has long lain outside of mainstream archaeological and historical discussions, I will also argue

39

40


that early (and often incorrect) ideas about the Asian state have influenced archaeologists’ expectations about the organization of the world’s earliest state societies, particularly as concerns interpretations of their managerial economic functions and high investment in economic domains, including craft production. This overview of the history of South Asian historiography is not intended as a comprehensive account of what is now a massive literature (see Cohn 1996; Guha 1988; Inden 1990; Kulke 1995a). I will emphasize only one of many diverse and contradictory strands of writings on the Asian state (see Teltscher 1995). The strand I trace concerns the political, and presents the dominant view of the Asian state that came to characterize European political and economic writings of the seventeenth through nineteenth centuries, and which in turn heavily impacted Asian historiography and archaeology. Such writings – by social theorists such as Adam Smith, James Mill, Karl Marx, Friedrich Engels, and Max Weber, among many others – were, more often than not, focused on Europe and its potential for political and economic change. In this context, concepts of the Asian state served largely as foils or negative examples by which the west could better view itself (Said 1978; Breckenridge and van der Veer 1993; Rubiés 2000).

T H E O R I E N TA L D E S P OT In his massive treatise Politics written in the fourth century BC, Aristotle defined the distinctive features of the Greek city-state, which he conceived as the natural and ideal political form (Reeve 1998: xlviii). He distinguished this ideal from two other kinds of states: illegitimate tyrannies, such as had occasionally arisen in Greece, and the legitimate but despotic states of Asia: But there is another kind of monarchy besides this, which is like kingships that exist among some non-Greeks. The powers all these have are very like those tyrants have, but they are based on law and are hereditary. Because non-Greeks are by nature more slavish in their character than Greeks, those in Asia being more so than those in Europe, they tolerate rule by a master without any complaint. So for this sort of reason these kingships are tyrannical, but they are stable because they are hereditary and based on law. (Book III, Chapter 14, 1285a: 16–24; 1998: 92)

These few sentences present several features that would come to characterize later views of the Asian state: the essential opposition between Europe and Asia, the tyrannical or despotic king, and his servile subjects (O’Leary


1989: 43). These traits are the source of the further defining features of the Asian state noted by Aristotle: its long-term stability, and the absence of any mechanisms for political change. Aristotle had little to say about the economic features of the Asian state. These were added in later formulations, following the revival of his writings in Western Europe in the thirteenth century (Sawer 1977: 6; O’Leary 1989: 46). There, the concept of the Asian despot re-emerged as a way of thinking about kinds of polities and leaders, now perceived in terms of the opposition between Christianity and Islam. Over the next three centuries, several additional features were added to the characterization of the Asian state. These included the belief that the Asian despot owned all economic resources, most importantly land: private property did not and could not exist. In addition, it was believed that Asian states were characterized by the absence of an autonomous aristocracy; subsidiary elites were seen as holding their positions at the whim of the ruler, rather than, as was the case in Europe, through hereditary status or access to independent sources of wealth (e.g., Machiavelli, 1988 [1513]: 17–19). With the expansion of European merchants and settlers into India in the sixteenth and seventeenth centuries, writings on the subcontinent became widely available in Europe, and India came to be viewed as the quintessential Asian state. The earliest of these writings were travelers’ accounts, later followed by reports of the East India Company and the emerging colonial authorities (largely British, but also French and Dutch; Teltscher 1995). Of particular relevance for Vijayanagara are the writings of two sixteenth-century Portuguese merchants who spent considerable time at the Vijayanagara imperial capital – Domingo Paes and Fernao Nuniz – (Sewell 1900; Rubiés 2000; see chapter 5). The Portuguese and their experiences in South India had, however, little impact on broader developments in political thought within Europe (though see Rubiés 2000). Here, pride of place must be given to the writings of two seventeenth-century French authors, Francois Bernier and Jean Baptiste Tavernier, whose works were widely read and provided primary source material for numerous later scholars, including Karl Marx. Both authors had resided in the court of the Mughal emperor Shah Jahan – Bernier from 1656 through 1658, and Tavernier from 1640 through 1666. Although somewhat different in emphasis, both stressed the Mughal emperor’s absolute control over land and people. Like earlier writers, they emphasized the absence of private property and the lack of a hereditary landed nobility as defining features of the Mughal state. Bernier provided

41

42


vivid descriptions of the wealth of the ruler and his court – including the fine silk and brocade garments, silk carpets, elaborate jewels and golden ornaments, the products of talented artisans. He noted, however, that such goods were never seen in the streets and markets of the empires, since they were only available to members of the court. Bernier devoted considerable attention to the economic limitations of the Mughal system (and its shortcomings compared to France), especially for the development of complex systems of non-agricultural artisan production. He argued that the complete dependence of artisans on the king, his nobles, and generals, rather than on broader communities of consumers, made technological progress impossible (Bernier 1968 [1891]; also O’Leary 1989: 10–11; Sawer 1977: 11). According to Bernier, the Mughal king granted some of his property in lieu of salary to his governors and military leaders, in exchange for troops and rent. These governors “have an authority almost absolute over the peasantry, and nearly as much over the artisans and merchants of the towns and villages within their district” (Bernier 1968 [1891]: 225). The resulting “debasing state of slavery obstructs the progress of trade and influences the manners and the mode of life of every individual” (Bernier 1968 [1891]: 225). Describing Delhi, Bernier observed “workshops, occupied by skilful artisans, would be vainly sought for in Delhi, which has very little to boast of in that respect” (Bernier 1968 [1891]: 254). This is because artisans were compelled by lords and generals to labor for them, under threat of force and for inadequate remuneration. As a result, they could not work independently and had no motivation to improve their art (Bernier 1968 [1891]: 254). The only place where the arts flourished was in the palace, where attached specialists produced sumptuous goods: Large halls are seen in many places, called Kar-kanays, or workshops for the artisans. In one hall embroiderers are busily employed, superintended by a master. In another you see the goldsmiths; in a third, painters; in a fourth, varnishers in lacquer-work; in a fifth, joiners, turners, tailors, and shoemakers; in a sixth, manufactures of silk, brocade, and those fine muslins of which are made turbans. . . . The artisans repair every morning to their respective Kar-kanays, where they remain employed for the whole day; and in the evening return to their homes. In this quiet and regular manner, their time glides away; no one aspiring after any improvement in the condition of life wherein he happens to be born. (Bernier 1968

[1891]: 258–259)

Bernier and Tavernier’s writings provided substantive “evidence” for the long-standing idea that Asian states were characterized by the absence of


private property, which in turn accounted for their lack of economic development and the impossibility of dynamic change. Bernier’s focus on the centrality of the king and royal household on artisan production would resurface in Max Weber’s later writings on patrimonial states. The eighteenth-century French author Montesquieu relied on Bernier and Tavernier in formalizing his concept of the oriental despot (Montesquieu 1989 [1749]). Montesquieu drew on missionary and travel literature to affirm the absence of property rights and hereditary nobility in despotic states. He added to this suggestions for explanatory factors – of geography and climate – that could account for the emergence and persistence of such states. For Montesquieu, despotism was a defining feature of territorially large empires (and hence not exclusively Asian). What was distinctively Asian, in Montesquieu’s view, was the contribution of Asia’s tropical climate, particularly as manifested in India, in contributing to the “natural indolence” of the region’s inhabitants. The innate torpor of the Asian subject made it possible for a ruler to “destroy the spirit of property” among his subordinates. Religion, Montesquieu believed, also contributed to the persistence of such polities. Without Christianity, resistance to these natural constraints could not occur and change was impossible. Montesquieu was not without his critics, including Voltaire, who praised oriental despotism (especially as manifest in China) and saw an enlightened despotism as a possible model for Europe (Sawer 1977: 17). However, his work was influential, and his environmental determinism would resurface in the writings of Engels in the nineteenth century and in the twentieth-century work of Karl Wittfogel. By the late eighteenth and early nineteenth centuries, a new modernist teleological view of history had emerged in parts of Europe, and human history came to be viewed in the west as “a dynamic progress of selfdevelopment” (Sawer 1977: 25). In this evolutionary conception of human (specifically northwest Europe’s) progress from savagery to civilization, Asian states were placed firmly as relics of an earlier form of political organization that European progress had long since superseded. This view was espoused by Georg W.F. Hegel, in his posthumously published Philosophy of History (Hegel 1956 [1837]). Hegel viewed the world historical sequence as progressing from ignorance and savagery toward freedom and reason. He argued that this process had bypassed Asia, which was stranded in the past and incapable of change (Sawer 1977: 26–28; O’Leary 1989: 69–70). For Hegel, Asia was the site of humanity’s earliest moves toward statehood, the place where history began. Its precocious forward movement had been halted by the despotism of Asian rulers and the absence of systems of

43

44


morality or subjectivity that could permit greater progress. Hegel reserved special enmity for India (Inden 1990: 51; O’Leary 1989), which he distinguished from China and Persia: The spread of Indian culture is prehistorical, for History is limited to that which makes an essential epoch in the development of Spirit. On the whole the diffusion of Indian culture is only a dumb, deedless expansion; that is, it presents no political action. The people of India have achieved no foreign conquests, but have been on every occasion vanquished themselves. (Hegel 1956 [1837]: 142)

E C O N O M I C M O D E L S : T H E A S I AT I C M O D E O F P RO D U C T I O N Oriental despotism entered intellectual discourse as a philosophical contrast to ideas about the nature of the European state, and, by the early nineteenth century, its basic features were well established and widely accepted among European intellectuals. In Britain, by this time heavily invested in South Asia, political economists expanded upon the economic dimensions of the Asian state. Thus, the eighteenth-century writer Adam Smith argued (1991 [1776]) that the distinctive feature of the Asian political economy was that the state (that is, the king) derived its wealth from tax or rent on land, rather than from commerce and manufacture (see also Bailey and Llobera 1981). Smith suggested that such a system had the beneficial effects of encouraging state investment in irrigation facilities and roads to transport agricultural produce, public projects not then engaged in by European states. But, on the negative side, the Asian system was inherently limiting, and did not allow for the emergence of a free market economy with potential for development and expansion. Half a century later, James Mill had nothing positive to say about Asian economies. Mill was a proponent of Utilitarianism, a radical philosophical school that argued for progress through modernization based on the goal of attaining the greatest good for the greatest number (Trautmann 1997: 117). Although a staunch critic of the growing conservatism of British political life (Majeed 1992), Mill agreed with his conservative foes that British domination provided the only route via which India could hope to reach the status of civilization. In 1817 Mill published his monumental The History of British India (Mill 1975[1817]). Despite its focus on India, Mill wrote his work as a critique of the traditional British legal system and of British practices in India and in Britain itself, and as an illustration of his Utilitarian principles


(Majeed 1992: 132–133). Mill sought to shift the British imperial project in India from a focus on profit to one of bringing civilization and progress to a benighted land. Although no longer considered an important example of Mill’s political and philosophical writings (in large part because of his vitriol toward India), the History had a major impact at the time of its publication, becoming the standard text for officers of the British East India Company and, later, a required book for candidates for the Indian Civil Service (Majeed 1992: 128). It was also severely criticized by many of Mill’s contemporaries, particularly by Orientalists and others with greater sympathies toward India and/or antipathies toward British imperialism. In his History, Mill presented a progressive view of historical change that discounted presumed past accomplishments in favor of future progress. This was accomplished in large part through a critique of the British Orientalists, particularly Sir William Jones, who had first identified the Indo-European language family that includes Sanskrit as well as Latin and Greek. Mill argued against the Orientalist view that exalted India’s ancient past as a time of former glories and high civilization, and that saw in India’s past the past of the west. For Mill, nothing noble could be learned from the study of ancient India. Having never traveled to India, Mill also sought to undermine the authority of the Orientalists’ India experience and firsthand knowledge. He argued that it was the metropolitan observer, unbiased by the confusing inputs of direct experience, who could best synthesize the scattered writings of diverse observers and provide the most authoritative account of India (Trautmann 1997; Majeed 1992). Like his predecessors, Mill (1975 [1817]: 35) portrayed India as an unchanging land: From the scattered hints, contained in the writings of the Greeks, the conclusion has been drawn, that the Hindus, at the time of Alexander’s invasion, were in a state of manner, society and knowledge exactly the same with that in which they were discovered by the nations of modern Europe . . .

Rulers were absolute and “possessed the whole power of the sovereign, to levy taxes, to raise and command troops, and decide upon the lives and property of the subjects” (Mill 1975 [1817]: 59). Mill’s significant new addition to this conventional knowledge was his emphasis on the village as the essential settlement form of India (see Inden 1990: 131–161). The village appears in Mill’s writings as an autonomous and self-sufficient “republic,” whose inhabitants plied their traditional lifeways independent of larger-scale political events:

45

46

The Political Economy of Craft Production The inhabitants give themselves no trouble about the breaking up and division of kingdoms; while the village remains entire, they care not to what power it is transferred or to what sovereign it devolves; its internal economy remains unchanged . . . (Mill 1975 [1817]: 126–127)

Here, we see merged the two seemingly contradictory major components – the “strange hermaphrodite” (Kulke 1995a: 4) – that would continue to dominate political and historical thought on South Asia, and would later resurface in the writings of Marx, Weber, and in nineteenth- and early twentieth-century ethnography on the Indian village. These were, first, the persistent idea of the despotic tyrant who controlled all land, resources, production, and subjects; and, second, the addition of the autonomous isolated village that persisted unchanged no matter what the despots did. Both components echo Aristotle’s mention of tyrannical kings and servile subjects written more than two millennia earlier. In the early 1850s, Karl Marx became interested in characterizing nonEuropean states, as part of his understanding of the trajectory of world history. Through his reading of Bernier, Mill, and another Utilitarian author, Richard Jones, Marx became convinced that the king’s ownership of all land was the essential feature of the Asian state. In a letter to Friedrich Engels dated June 2, 1853, Marx wrote that “Bernier correctly discovers the basic form of all phenomena in the East – he refers to Turkey, Persia, Hindostan – to be the absence of private property in land” (emphasis in original, reprinted in Sawer 1977: 43). In his reply, Engels echoed Montesquieu in suggesting that the reasons for this lay in geography, though for Engels the main factor was the necessity for large-scale irrigation works requiring state management. The dominant village mode of settlement of the region contributed to the power of the state, as isolated self-sufficient villages whose inhabitants engaged in agriculture and craft production (particularly weaving) were unable to collaborate independently in the large-scale enterprises necessary for agriculture. Like Tavernier and Bernier, Marx believed that true urbanism, with its possibilities for independent commerce and manufacturing activities, did not exist in India as it did in Europe. Instead, the ruler and palace directly controlled all production of luxuries, while production of ordinary goods, such as textiles, occurred in the isolated villages where the majority of the Indian population lived in unchanging isolation. These two circumstances – the Hindoo, on the one hand, leaving like all Oriental peoples, to the central government the care of the great public works, the prime condition of his agriculture and commerce, dispersed, on the other hand, over the surface of the country, and agglomerated into small centers by the domestic union of

The South Asian state agricultural and manufacturing pursuits – these two circumstances brought about, since the remotest times, the so-called village system, which gave to each of these small unions their independent organization and distinct life. (Marx 1968 [1853]:

86–87)

Marx and Engels’ model of the Asiatic Mode of Production placed the Asian state as an early form of state that, having emerged, was incapable of change. We must not forget that these idyllic village communities, inoffensive though they may appear, had always been the solid foundation of Oriental despotism, that they restrained the human mind within the smallest possible compass, making it the unresisting tool of superstition, enslaving it beneath traditional rules, depriving it of all grandeur and historical energies. (Marx 1968 [1853]: 88)

Class-consciousness could not arise among India’s ignorant and isolated peasants and rural artisans, and therefore there could be no possibility for villagers to act corporately to alter or improve the conditions of their own existence. Numerous lengthy studies of the Marxist concept of the Asiatic Mode of Production have appeared (see, Bailey and Llobera 1981; Krader 1975; O’Leary 1989; Sawer 1977) and the concept has been subject to criticism and modification (Anderson 1974). It was discredited by many later Marxist thinkers in South Asia and elsewhere and, in India, has been largely replaced by the concept of Indian feudalism (see discussion later this chapter). For archaeology, the most significant expansion of Marx and Engels’ Asiatic Mode of Production came from the hydraulic theory of Karl Wittfogel, presented in his 1957 book Oriental Despotism. Like Engels, Wittfogel emphasized Asia’s distinctive environmental conditions, particularly the dominance of large river systems in semiarid environments. Under such conditions, surplus agricultural production could only come about through hydraulic agriculture based in large-scale irrigation works, and such projects and the massive labor investment they demanded were only possible under the authority of highly centralized despotic state administrations. Oriental despotism was the inevitable result in contexts where state investment in agriculture was a necessary precondition to surplus production. Wittfogel argued that the hydraulic state reached its fullest form in China and India (although it was not restricted to Asia, being also characteristic of the Andean Inkas). The heavy labor investment in constructing and managing hydraulic works could only be accomplished through a centralized despotism and state control of all labor and resources. The major characteristic of the despotic hydraulic state was a centralized bureaucracy, focused

47

48


on the absolute power of the king who controlled the labor of “villagebound” or free (non-slave) peasants (Witffogel 1981 [1968]). Wittfogel did distinguish among different categories of oriental societies, which he termed simple and developed. In simple despotisms, “the sovereign and his bureaucracy exercise material control over the totality of cultivated lands” (1981: 150) and the consequent surplus produced. In developed oriental societies, private property exists to a limited extent, though in the non-agricultural realm of commodity production and trade (1981 [1968]: 150–151). However, in either instance, Wittfogel shared with earlier scholars the view that Oriental despotisms were a highly stable political form, incapable of structural change. Asiatic society, once established, was a “stationary society” (Wittfogel 1981 [1968]: 157): it could repeatedly reproduce itself, but could generate nothing new.

E A R LY H I S TO R I O G R A P H Y : D I S C U S S I O N As outlined above, the distinctive features of the Asian state have an extraordinarily long history in western political thought, and can be traced from the fourth century BC until the twentieth century AD. This history was certainly never entirely uncontested or lacking in alternative presentations. Teltscher (1995) has argued that, from as early as the sixteenth and seventeenth-centuries, accounts of India appeared in Europe that questioned those of Bernier and Tavernier. However, the view of the oriental despot remained dominant, in political theory and in practical application in legitimating European colonial intervention in the region. And its impact on both public perceptions and on historical and archaeological research has been profound. Ideas of a timeless unchanging East remain pervasive in western (and Indian) popular thought, and in many academic works. Mill’s concept of village India spawned generations of ethnographic research and supported the concept of a political stability (stagnancy?) that could only be impacted by foreign migration. These concepts remain powerful in the archaeological literature of both prehistoric and historic South Asia. Craft production is not a major concern in these writings, as authors consistently emphasized land and agriculture. Nonetheless, references to commodity production appear in the writings of Marx, who built on Bernier, Tavernier, Mills, and others. For these authors, craft production occurred in only two contexts: isolated village communities or attached to royal palaces. In either setting, producers exercised no control over their lives or products; they were trapped either in the “undignified, stagnatory and vegetative life”


(Marx 1968 [1853]: 89) of the village or in dependency on the despot. Classconsciousness and class struggle were not possible in such circumstances. The state’s ownership of all critical resources precluded the development of private property, large-scale commodity production, independent entrepreneurs, or a vibrant mercantile economy. Innovation and creativity were stifled, and economic or historical progress was thus impossible. Takings these ideas beyond South Asia to the “Orient” more broadly, the archaeologist V. Gordon Childe applied this framework to his writing on the early Near East, including the Hegelian emphasis on the Orient’s stagnancy compared to Europe (e.g., Childe 1925, 1952; Trigger 1980). And, as discussed above, Wittfogel (1957, 1981 [1968]) applied the Asiatic state model to the ancient world. Although largely discounted in its particulars, particularly on the role played by irrigation in state economies, Wittfogel was successful in defining the main role of the ancient state as managerial, and concerned with the efficient production and control of essential resources. This view of the early state received further boosts in the 1960s and 1970s, when it intersected with new theoretical approaches in archaeology drawn from general systems theory and adaptionist approaches to cultural evolution (e.g., Flannery 1972; Wright and Johnson 1975; see discussion in Blanton 1998). However, unlike Marxist-influenced approaches that focused on the repressive and negative characteristics of such polities, the adaptationists emphasized their positive effects in fostering the coordination of large numbers of people and resources through centralized authority and decision-making structures. The legacies of Orientalist models of the state, while enduring, have been extensively critiqued both within South Asian history and, less directly (since this legacy is largely unacknowledged), within archaeology. Over the last two decades, both fields have moved toward views of Asian and other states as more dynamic and less effectively centralized than any of the authors discussed above allowed (see G. Stein 2001, for a recent discussion of the archaeological literature on early states), and as the Vijayanagara evidence certainly supports – an issue I will address in more detail in chapter 8. In the remainder of this chapter, I shift focus to more recent writings on the South Asian state, in which several distinctive approaches have been debated, and where South India and Vijayanagara have figured prominently. Some recent approaches, such as the concept of Indian feudalism, are a direct response to and critique of earlier models. Others, such as the segmentary state model of Burton Stein, entail radical refutations of the idea of the despotic state, though Stein’s original formulation retains the implicit idea of a distinctive Asian “other” that lacks the dynamism characteristic of state societies in other areas of the world.

49

50


T H E H I S TO R I C S TAT E I N S O U T H A S I A : R E C E N T A P P ROAC H E S Hermann Kulke (1995a) has recently summarized five distinct models in approaches to the state in precolonial India, specifically for the period from AD 1000 to 1700. These are: 1. The Marxist Model of Oriental Despotism and the Asiatic Mode of Production 2. The Indian Historiographical Model 3. Indian Feudalism 4. Segmentary State Model 5. Patrimonial State Model Kulke adds to this a sixth category, in which he places recent works by Chattopadhyaya, Heitzman, and himself that do not conform to any of the existing universalizing models and focus instead on the historical trajectories of individual polities or regions. Since I have discussed oriental despotism and the Asiatic Mode of Production earlier, I only note here that they seldom appear in contemporary literature as explicit formulations. However, as I have stressed above, many of their underlying assumptions continue to figure in significant ways, and they are present or responded to in one form or another in several of the other approaches summarized by Kulke. Thus, many supporters of the Indian feudalism approach are responding to Marx’s Asiatic Mode of Production model in an effort to shift South Asia from a blind alley on the world historic trajectory to a more dynamic position. For Vijayanagara, a feudal interpretation is most explicit in the writings of Noburu Karashima (e.g., 1992). I will turn to specific interpretations of Vijayanagara in chapter 4. Here I briefly examine the underlying tenets and assumptions of the “Indian historiographical,” feudal, segmentary state, and patrimonial approaches to the South Asian state.

The Indian historiographical model What Kulke calls the “Indian historiographical model” has been termed the “conventional” model by Burton Stein and the “imperial” model by Shulman (Kulke 1995a: 4). This approach both derives from and contains within it radical critiques of views of the state widespread in British


colonial writings on India’s precolonial history. Manifest most explicitly in the writings of historian Vincent A.S. Smith (1958 [1919]), the latter were primarily political histories focused on the grand sweep of a long succession of historic states and dynasties. Smith and his contemporaries posited the very early existence of highly centralized and bureaucratic imperial states during the Mauryan and Gupta periods (pre-sixth century AD), followed by a “medieval” period of fragmentation and stagnation that persisted until the arrival of Central Asian Turkic-speaking invaders in the twelfth century. Smith shared with other proponents of oriental despotism a sense of India’s precociousness in its early history of statehood and of its later decline. What is absent is any consideration of underlying economic or other factors that could have led to such polities. Smith, like Hegel and Mill, also stressed India’s inherent disorder and tendency toward dissolution and corruption, and the consequent need for British imperial domination (see Inden 1990: 8–9). Proponents of the Indian historiographic tradition hold a sharply different view of Indian history. They share Smith’s chronology of early state centralization followed by an extended interregnum of decline and fragmentation. However, they typically saw a resurgence of centralization in the later “medieval” period, under the Cholas and Vijayanagara in the south and the Mughals in the north. Leading scholars of this tradition, such as R.C. Majumdar and K.A. Nilakanta Sastri (Kulke 1995b), also agreed that ancient Indian states were characterized by strong despotic rulers and highly centralized bureaucracies bolstered by a monopoly of coercive force. However, proponents of the Indian historiographic tradition recast the image of coercive state power of the despotic model in a positive light, as evidence for the existence of strong, effective, and highly centralized states in precolonial contexts. Sources cited to support this perspective for early India include ancient texts such as the Arthasastra and the Ashokan inscriptions, both of which pose considerable interpretive problems (see Trautmann 1971; Sinopoli 2001b; Thapar 1997). This image of powerful and legitimate historic Indian states became prominent among Indian historians of the early to mid-twentieth century and figured importantly in the Indian freedom struggle. Jawaharlal Nehru, independent India’s first prime minister, made frequent reference to India’s precolonial unity and stability under the Mauryans as evidence for India’s capacity for independent governance (Kulke 1995a: 4; Sinopoli 2001b). Despite the empirical problems in the interpretations of Mauryan and Early Historic India political structures, the Indian historiographical tradition remains the most widespread and least theorized approach in Indian

51

52


historical writings. It has contributed to a highly problematic periodization of Indian history – into Buddhist, Hindu, and Muslim periods, and their respective golden ages – in ways that dangerously simplify the complexity of political and religious relations in the region, and have major ramifications in contemporary ideological and political debates (Thapar 1989). While not explicitly linked to the concept of oriental despotism, the approach undoubtedly derives from it, and shares a concept of extremely strong rulers with dependent managerial bureaucracies, as well as an emphasis on isolated and stable village communities. It also emphasizes India’s precocious early development of highly centralized states, followed by a later instability (though differs in acknowledging subsequent periods of recentralization before British colonialism).

Feudalism in South Asia As previously noted, scholars of the Indian historiographical tradition present a trajectory of the Indian state that begins with the early highly centralized states of the Mauryan and Gupta dynasties that, it is argued, were followed by an extended period of decline and fragmentation, lasting from c. AD 400 to AD 1100/1200. A similar sequence is propounded by proponents of quite a different approach to Indian history – Indian feudalism (see also chapter 4, pp. 114–116). For both groups, the postGupta “medieval” period (Chattopadhyaya 1994; Sahu 1997) was, like the coterminous period in Europe, a time of economic and political contraction. Specifically, medieval India is seen as marked by political fragmentation, rural isolationism, the disappearance or decline of urbanism, and the collapse of long-distance trade networks and monetary systems (see Sahu 1997: 4). The theoretical and historiographical challenge is in accounting for this transition from early large, centralized, imperial states to a “Dark Age” of small, closed, and isolated agrarian polities. The medieval or “middle” period was not a significant focus of early research on historic South Asia. This changed in the 1950s with the work of several prominent Marxist historians, motivated by an interest in economic models and a desire to provide a Marxist refutation for the Asiatic Mode of Production model (Kulke 1995a: 7; Anderson 1974). They found this refutation in the concept of a distinctively Indian feudalism. Since then, arguments over whether or not feudalism existed in historic South Asia have been heated (see Mukhia 1995, 2000; Jha 2000; Chattopadhyaya 1994), and Kulke (1995a: 6) has called this “the most controversial issue of modern


Indian historiography.” The literature for and against Indian feudalism is by now extensive, and I do not attempt a full review here. Supporters of the Indian feudalism perspective (e.g., Kosambi 1956a, 1956b; Sharma 1965, 1987, 1995; Jha 1987; Yadava 1997; Thakur 1989) differ in emphases and in the specific criteria and sources of evidence they use for identifying the feudal mode of production in India and accounting for its emergence. All stress that, although they draw on European models, Indian feudalism differed in significant ways from that of Europe. However, like European medievalists, the Indian feudal perspectivists argue that feudalism emerged out of the collapse and fragmentation of earlier centralized states (although not necessarily “slave society”; however, see discussion of Karashima in chapter 4), and the reallocation of property into the control of a restricted elite segment of the population. Coupled with a decline in urbanism and state weakness, the result was, to quote Goldstone (1998: 266–267), “decentralized congeries of ‘lords’ depending on control of land and labor, with limited urban centers and manor-centered economies, and no bureaucratized and centralized government collecting taxes in recognized national currency.” Proponents also disagree about the pervasiveness of Indian feudalism. Thus, Yadava (1997: 336) argued that feudal traits coexisted with non-feudal features, and that there was somewhat greater economic and social mobility among Indian peasants than European serfs. The primary historical evidence used to argue for feudalism comes from inscriptions and texts that record the transfer of land by rulers to local elites, including Brahmans and religious institutions, as well as local leaders and military officers. Thus, it is argued that land was taken out of the direct control of the state and peasant laborers, who are believed to have held property rights in pre-medieval periods, contrary to the expectations of the Asiatic Mode of Production view. A new class of elites was created, with juridical and economic control over land, revenues, and labor. Brahmadeyas, or the lands granted to Brahmans, and land grants are indeed known from numerous temple inscriptions in many regions of India, and, it is argued, provide evidence for diminution of peasants’ rights to land and its products (e.g., Sharma 1995: 52). Brahmans and other landholders in turn became intermediaries between the weakened state and producers (though differences exist in interpretations of whether they had reciprocal service obligations to the crown such as the kind the European feudal lords had to their king; see Kulke 1995a). Critics of the feudalism approach argue that although these brahmadeya settlements are documented in the inscriptional record, they were by no means the only form of land ownership or agrarian production that existed

53

54


(Sahu 1997: 29; also Champakalakshmi 1995: 277), and that we need to consider the biases of the inscriptional record and their overemphasis on temples and Brahmans. Nor did a single trajectory characterize all of South Asia during the medieval period. Instead, brahmadeyas coexisted with various kinds of peasant landholders, royal estates, sharecroppers, landless laborers, and slaves, creating a much more complex system of agrarian production and social relations than could be glossed under a concept like feudalism. Thus, while feudal relations may have characterized some landholders, producers, and landholdings, feudalism as a system was not exclusive, nor even dominant. Even for the brahmadeyas, debate exists over the extent to which peasants were tied to the land or to the Brahman landowners. Mukhia (1995: 127) has argued that “free peasant production” (in which peasants controlled their own labor) predominated throughout the medieval period. Chattopadhyaya (1986, 1994, 1995) and others (e.g., Champakalaskhmi 1991, 1995; Chandra 1991: 83–84) have argued that the concept of a post-Early Historic disintegration is greatly exaggerated, in its arguments for a decline both in trade and in urbanism. They argue instead for approaches that address temporal continuity and regional variability, rather than trying to construct a single frame for all of the “medieval,” or any other period, of South Asian history. Little has been written explicitly about the impact of feudalism on craft production. However, the arguments for the disappearance of powerful states, urbanism, and long-distance exchange, and the expectation of an isolated and dependent peasantry, all imply a decline in the role and power of merchants and artisans and an overall decrease in production. Devangana Desai (2000) has argued that the disposition of wealth in the hands of Brahmans, temple elites, and other elite landholders, encouraged the development of certain crafts dedicated or attached to elite patrons, including temple architecture, sculpture, and laudatory songs and poetry. For Desai, the result was uncreative and monotonous, especially as compared to earlier more dynamic periods; the resultant art displays “the ossification of form and spirit” of production in feudal contexts (Desai 2000: 493). Vijayanagara falls at the end of the traditional chronology of the South Asian “medieval” period, and most scholars of the Indian feudal tradition have typically stopped their discussion before the Vijayanagara period. However, historian Noburu Karashima (1992) has applied a rather distinctive feudal interpretation to Vijayanagara and South Indian history. In particular, he views the South Indian Chola period (ninth to twelfth centuries AD) as characterized by state slavery and traces the transition to feudalism only


to the late fifteenth century. I will consider Karashima’s view of Vijayanagara in chapter 4.

The segmentary state The concept of the segmentary state, developed by anthropologist Aidan Southall to describe the political organization of the African Alur society, was first applied to South Asia in the 1970s by the American scholars Richard Fox (1971) and Burton Stein (1977, 1980, 1985, 1989a, 1995). Stein’s work on the Chola (ninth to thirteenth century AD) and Vijayanagara (fourteenth to seventeenth century AD) empires provides the most elaborate and influential approach to the segmentary state concept, and his writings have been both lauded and heavily criticized (e.g., Champakalakshmi 1981; Chattopadhyaya 1995). From either perspective, Stein contributed to a dramatic expansion of research on South India, and, for the first time, brought the region to the forefront of theoretical discussions on Indian political structures. Stein opposed the segmentary state to the idea of centralized states characterized by fixed territories, centralized administration, and coercive power. Segmentary states, in contrast, lack formally defined territories and are characterized by numerous centers or domains. Each of these many centers has autonomous administrative and coercive authority, and some degree of political power. Each is also largely autonomous economically, and while resources do flow between hierarchical levels of the segmentary structure, in the form of taxes and tribute, such flows are often limited and/or mediated by temples. Rather than the political or economic, what molds these segments together into a single “state” is an acknowledgment of the sovereign authority, particularly sacred authority, of a single ritual center and legitimate king. Instead of subjugation by a despotic ruler, it is the widely shared concept of lordship or a quasi-sacred monarch that links the disparate and often competing political units into a self-acknowledged whole. Stein (1995: 143) conceived of South Indian society as “a generalized polity of chiefdoms . . . a system of enduring (but not necessarily permanent) political structures, based on strong hereditary regimes with extensive authority over wide areas and over varied and internally ranked local social segments.” For Stein (1995: 151), the segmentary state concept provides a formal structural model that explores the relations between components of a political structure. He did not link his framework to the anthropological concept

55

56


of segmentary kinship (of which he was unaware when he proposed his model), nor does his segmentary state require particular kinds of economic structures or relations. As such, issues of craft production and political economy play no role in his framework. Instead, what is essential is that distinct social segments, however defined, must exist and have historic continuity, and they must play an important role in organizational relations. But the ways such segments are constituted – by kinship, caste, territory, etc. – can vary from case to case. As a structural model, the segmentary state model lacks dynamism or recognition of mechanisms for significant structural change. Stein did acknowledge change in South India, particularly in his discussion of the Vijayanagara period and of various attempts by Vijayanagara rulers at greater state centralization (1989a, 1995; see chapter 4). Nonetheless, he believed that the segmentary structure was both highly durable and accurately characterized Chola and Vijayanagara political structures throughout their centuries-long histories (a view that has been much criticized by Champakalaksmi 1981; Heitzman 1991, 1995, 1997; Palat 1987; and Karashima 1992). Stein first developed his segmentary state model in a study of the Chola state of ninth- through twelfth-century Tamil Nadu (B. Stein 1977, 1980). He argued that the relevant segments or units for the Chola were territorial units, localities or nadu, which nonetheless varied somewhat in constitution and in their interconnections in central, intermediate, and peripheral areas of the state. His views of Vijayanagara evolved over the more than two decades that he wrote on Vijayanagara state organization. Vijayanagara appeared in his 1980 book as a second example of a segmentary state, comparable to the Cholas. In 1985, Stein wrote of Vijayanagara as transitional between segmentary and “patrimonial” forms of state organization, and he viewed Vijayanagara as “progressing” from a segmentary to a more centralized state form. This emphasis continued in his 1989 book on Vijayanagara (B. Stein 1989a); Stein still considered Vijayanagara a segmentary state, but the concept was no longer at the center of his discussions of the economic and political dynamics of the empire. It re-emerged at the center of Stein’s 1995 response to his critics, in which Vijayanagara appeared again as an important exemplar of the segmentary state, although different in form than the Chola state. In contrast to the traditional territorial segments of the Cholas, the foci of Vijayanagara’s political and economic segments were military leaders, who were ceded territories in exchange for service to the crown, and coastal chiefs, who benefited from expanded long-distance trade, particularly in textiles (bringing at least this sphere of craft production into the


mix). Further, despite expanding militarization, the persistence of a variety of communal entitlements among diverse regional groups (e.g., religious communities, caste groupings, artisan communities, etc.) also limited the ability of Vijayanagara to exert effective centralized control over its territories (B. Stein 1995: 149–150). I will address Stein’s specific interpretations of Vijayanagara in more detail in chapter 4. Of greater significance here, Stein’s major contribution to the study of the state in South Asia was in energizing research on the political organization of southern India. Stein brought a previously under-theorized region (Subrahmanyam 1998c) into the center of intellectual debates on the nature of the South Asian state. He provided scholars a new and synthetic way of thinking about political organization and its relations to economic and ideological formations and processes. However, Stein’s model, like other models of the South Asian state, provided no means for addressing the question of temporal or structural changes in state organization. Stein’s views evolved as a response to the Indian historiographical model that viewed the later Chola and Vijayanagara states as highly centralized and bureaucratized. Unlike Nilakanta Sastri, Stein did not see evidence for such centralization and called upon Southall’s segmentary state model to explain the region’s deviations from the norms of traditional centralized states. In this, Stein shared with other scholars a sense of the (South) Indian state as different from other kinds of states. In his case, this state was characterized by ritual rather than political/administrative sovereignty, comparatively weak links among its parts, and high local autonomy. The segmentary state could only be understood as opposite to, or the negative of, a normative centralized state. And Stein’s view of the latter resonates with the concepts familiar to us from earlier discussions – centralized states are characterized by a high degree of centralized political authority, well-developed bureaucracies, effective and pervasive revenue collection systems, and absolute control of force or power. I agree with Stein, that Vijayanagara did not conform to this conception of a centralized state. I argue, however (see Sinopoli 2001a; chapter 8, this volume), that in extensive imperial states, such as the Chola or Vijayanagara (or even the Romans or Inka), it is the highly centralized organizational form that is aberrant or perhaps non-existent. In imperial states, political structures entailing diffused systems of authority, weak territorial boundaries, and multiple centers of power are the norm, and not the exception (see Sinopoli 1994a, 2001a). By setting up a contrast with a kind of state that rarely if ever existed in human history, Stein’s segmentary state model for South Indian states leads us in a direction that continues to isolate South Asia

57

58


from broader considerations of the working, consequences, and histories of states and empires.

The patrimonial state In 1985, Stein wrote of Vijayanagara as transitional to a patrimonial regime. The concept of the patrimonial, or patrimonial-bureaucratic, state, derived from the writings of Max Weber, has been most extensively applied in South Asia to the Mughal empire by historian Stephen Blake (1980, 1986, 1990, also Richards 1993). Weber (1978: 231) viewed the patrimonial state as among the “most elementary types of traditional domination.” Patrimonialism and, in the extreme, sultanism tend to arise whenever traditional domination develops an administration and a military force which are purely personal instruments of the master. Only then are the group members treated as subjects. (Weber 1978: 231–232)

Patrimonial systems are patriarchical households writ large, with administrators and military officers serving as “patrimonial retainers” of the ruler (Weber 1978: 235). According to Weber, these retainers may receive support through five mechanisms: a. by living from the lord’s table, b. by allowances (usually in kind) from the lord’s magazines or treasury, c. by rights of land use in return for services (“service-land”), d. by the appropriation of property income, fees or taxes, [and] e. by fief. (Weber

1978: 235)

Stephen Blake’s study of Mughal imperial organization uses the Weberian patrimonial model as a frame for considering Mughal political structure, which he believes has been wrongly interpreted as a heavily bureaucratized and centralized polity by scholars of the Indian historiographical tradition (e.g., Ali 1966, 1972). The Mughals, for Blake, should be classed as a “patriarchical-bureaucratic” empire, a variant of the patriarchal state that is a response to state expansion in demographic and physical scale. In the patrimonial-bureaucratic state, rulers must institute additional means of maintaining loyalty when personal relations and face-to-face contact with retainers become difficult. These may include intelligence networks, periodic rotation of officials, and hostage taking, as well as the ruler’s high mobility across his territories, on tour, pilgrimage, or military missions (Blake 1995: 283; Sinopoli 1994b). Blake relies on the Mughal political text Ain I-Akbari (“Regulations of Akbar”) – composed by the minister Abu al-Fazl during


the reign of the emperor Akbar (AD 1556–1605) – for his reconstruction of the Mughal political system. Not surprisingly, Akbar figures at the center of the work as the ideal ruler and father of his people. Indeed, Blake (1995: 302) describes the empire of this period as “the state Akbar organized.” Key features described in the text include Akbar’s ultimate authority over the military, the imperial mint, revenue collectors, and various levels of officials and spheres of production, centered in large workshops – the Kar-kanays, described by Bernier. Blake’s patrimonial-bureaucratic Mughal empire was a strong polity, but it lacked enduring political structures or stable administrative systems, since power and authority ultimately rested in the success and strength of the individual ruler, and efforts had to be continuously made to enforce the ruler’s potentially fragile position. While Blake acknowledged continuity between kings in political, economic, and military structures, much of the effectiveness of the state was vested in the distinctive qualities and legitimacy of individual rulers.

C O N C LU S I O N As noted earlier, Kulke’s summary of approaches to the state in “middle period” South Asia (AD 1000–1700), includes an “other” category that contains what he has termed “ ‘non-aligned’ studies.” These studies do not accept any of the models discussed above, and instead “focus on structural developments and changes within a given state system” (Kulke 1995a: 2). Included in this loosely defined category are works by Heitzman (1995, 1997) on the Chola state, Kulke (1995b) on Orissa, Chattopadhyaya (1994) on early medieval Rajasthan, and Talbot (2001) on the Kakatiyas of Andhra Pradesh, among others. Discussion of the specifics of these works falls beyond the aims of this chapter. Nonetheless, it is in such writings, many of which have emerged out of a dissatisfaction with the more universalizing approaches described above, that we see some of the most exciting developments in contemporary historiography. These include, first, an emphasis on analysis of detailed empirical evidence of a variety of sorts and from a variety of contexts (largely inscriptions and sacred and secular texts; evidence from archaeology remains rare) to elucidate the structures of individual polities. Thus, Heitzman’s work relies on Chola inscriptions from several ecologically and historically distinct regions to explore organizational variability over time and place. Such an emphasis on variability becomes especially important in the study of

59

60


large imperial polities, such as the Vijayanagara or Mughal empires, that, by definition, consisted of discontinuous territories and distinct subject populations with diverse histories, economies, belief systems, etc. (see Morrison 2001; Sinopoli and Morrison 1995). Many of these subject populations were not literate, nor were they necessarily written about. Archaeological data on such communities will ultimately provide a valuable source of evidence to explore their experiences of, and roles in, particular state structures. Second, several of these studies have focused explicitly on change over time in state institutions and relations – considering both structures and mechanisms for change. They make explicit the recognition that historic South Asian states were not timeless, but instead, that political, economic, social, and ideological structures, and the lives of their subjects, could be and were transformed by human action. A third important feature of several of these works lies in their emphasis on indigenous concepts or political understanding for thinking and writing about historic state structures (Subrahmanyam 1986; for example, Chattopadhyaya 1994, 1995). Fourth, these studies are characterized by a reluctance to present or develop a universal model of state organization for South Asia. Each is restricted to particular times (or time spans) and regions of the subcontinent, and while not unconcerned with the broader implications for general theory building, none seeks the definitive model for “The South Asian State.” This is a most salutary development. It makes little sense to expect a single model of state organization in a region that encompasses both the extraordinary cultural, environmental, and historical diversity, and the sheer physical scale of South Asia (any more than we would expect a single model of state organization to apply to all of Europe for the last two millennia). Certainly, the many people and polities that have inhabited this vast region during the five millennia since states first emerged in the northern subcontinent have engaged in widespread interactions of a wide variety of sorts, and have created something that is distinctively South Asian. Yet a salient feature of this something, whether we call it culture, civilization, or by some other term, is that it is highly varied, as befits this vast and densely populated area. Unlike scholars of the seventeenth through mid-twentieth centuries, contemporary scholars (including proponents of segmentary and patrimonial approaches) no longer expect that a single type of state level political organization emerged or persisted across the region and across millennia. This acknowledgment of variability and recognition of a multiplicity of models for state organization may seem at odds with the argument for broad comparative and contextual perspectives that I presented at the start of this


chapter. However, it is not. While it is important to avoid homogenizing all of South Asian history into a single interpretive framework and to appreciate the particularities of polities and times, it is also valuable to consider broader patterns of commonality and difference within and beyond South Asia. During the Vijayanagara period, as in all periods of South Asian history, the region was not isolated from the rest of the Old World (and by the sixteenth century, the New World, as well). Traders, pilgrims, diplomats, and refugees moved widely by sea and land, and goods, knowledge, technologies, and belief systems flowed out of and into South Asia through many ports and routes. This is one important reason to look beyond the territorial bounds of the region. Another and ultimately more significant reason for comparativism in the study of the state lies in the fact that in many of the models of South Asian state formation that I have discussed here, there has been an implicit, and sometimes explicit, “other” kind of state to which the South Asian state has been contrasted. For Aristotle, this other was the Greek city-state, for Machiavelli, the Renaissance principality, for Monstesquieu, the French monarchy, for Mill, the British empire. In those writings, South Asian states were seen as less than or negatives of the ideal states of which their authors were citizens. The idea of other kinds of state to which South Asian states are opposed persists in the writings of more contemporary authors, most explicitly in Stein’s contrast of segmentary with centralized states. Stein has rejected the centralization of Oriental despotism and of the Indian historiographic approach as unsupported by the South Indian evidence, yet maintains the concept of an ideal “centralized state” that one would presume, had an empirical existence at some time and place (for Stein (1989), that place was fourteenth- to sixteenth-century western Europe; see also Rubiés 2000). I am not suggesting that it is inappropriate to have alternate scenarios in mind when considering a particular case – indeed, quite the contrary. However, it is important to recognize the historical contexts under which these alternative understandings of states developed, and to evaluate their limitations and legitimacy. The emphasis on variability in state organization that characterizes the new approaches to the historiography of South Asia, also characterizes recent approaches to the Roman empire, and to numerous other geographically extensive pre-modern states (Garnsey and Saller 1987; Sinopoli 1994a, 2001a; Alcock et al. 2001). As Chattopadhyaya has cogently written, “in no state system, however centralized, can there be a single focus or level of power, and the specificity of the differential distribution of

61

62


power . . . may be an issue more complex than has hitherto been assumed” (1995: 207; see also Ludden 1996). Recent scholarship in many areas of the world suggests that the understanding of this complexity is a key challenge for scholars of ancient states. Studies of South Asian states, derived from rich archaeological and textual sources, can both benefit from and contribute to these broader theoretical understandings with implications far beyond the geographic bounds of South Asia.

4

Vijayanagara: the historical setting

The rulers and institutions of Vijayanagara forged a polity that drew on a two-millennium long tradition of state societies in South Asia. Political forms, sacred institutions, complex systems of production and exchange, and social institutions such as guilds and caste organizations all had a lengthy history in southern India, and any study of Vijayanagara must take into account this historical framework. While pre-existing structures did not limit the possibility for change, they did constrain the forms that changes could take and the range of choices available to individual actors and institutions. Like other domains, specialized craft production also had a long history in South India before the emergence of Vijayanagara. Producers belonged to hereditary occupational groups – castes and subcastes – that also carried enormous social and ideological weight. Many of the relations among craft producers and between producers and state institutions, temples, and merchants that we see as characteristic of the Vijayanagara thus had their roots in earlier developments. In this chapter, I present an historical overview of Vijayanagara – its dynastic history, and the political, economic, ideological, and social structures relevant to the study of Vijayanagara-period craft economies. Throughout the history of the Vijayanagara polity, the territories that the state could claim to rule were far from stable, and often even areas that nominally acknowledged Vijayanagara sovereignty were largely autonomous of its authority. It is thus perhaps more appropriate to speak of a general temporal and geographic range (i.e., “fourteenth- through seventeenth-century South India”) rather than the “Vijayanagara period” or “Vijayanagara territories,” phrases I will continue to use for the sake of brevity.1 This is especially the case because many of the changes we see in craft production and consumption during the period emerged out of broader social, economic, and ideological trends that had complex causes, of which the Vijayanagara polity was only one component. Included among these trends, many of which had begun in pre-Vijayanagara periods, were: (1) an overall intensification of agricultural and craft production; (2) the growth and spread of urban centers;

64


(3) the growth of large temple centers sponsored by and a focus of competition among local and regional elites; (4) increased monetization of the economy and concomitant changes in taxation and revenue transfer; (5) dramatic demographic shifts, including the expansion of Telugu elites and agriculturalists into Tamil-speaking regions of southeast India and migrations to urban centers and to newly cleared forest zones; (6) the expansion of international and internal commerce; and (7) a restructuring of caste relations and social identities at both local and regional levels. In this chapter, I will look beyond the political to the broader conditions that characterized different regions of South India in the fourteenth through seventeenth centuries. I set the stage for this discussion by providing historical background on several major South Indian states that preceded Vijayanagara, and in which we begin to see the start of a number of the developments that would intensify during the Vijayanagara period. I next summarize Vijayanagara dynastic history, imperial geography, and political structures, before turning to discussions of several economic and social issues that have particular relevance to the study of craft production. I conclude the chapter with a discussion of the works of three major scholars who have presented quite different interpretations of Vijayanagara, with distinctive implications for the study of political economy: K.A. Nilakanta Sastri, Burton Stein, and Noburu Karashima.

H I S TO R I C A L C O N T E XT As noted, societies characterized by political hierarchy and economic specialization have a long history in South India. Structured relations of social inequality emerged in some regions of the South by the early first millennium BC, during what archaeologists refer to as the “Iron Age.” That period is best known for its large megalithic cemeteries, which provide evidence for dramatic differences in mortuary treatment and grave goods among individuals and between communities (Brubaker 2001; Moorti 1994). Relatively few Iron Age settlements have been well documented. Nonetheless, in several regions, there is evidence for settlement hierarchies, complex agricultural and pastoral economies, and economic and social differentiation. And as this period glides into the succeeding “Early Historic Period” (the boundaries between the two being difficult to distinguish and necessarily


arbitrary), textual sources from some areas provide a vivid picture of a complex social and economic hierarchy dominated by warrior elites (i.e., the Tamil Çangam literature of the late centuries BC; Champakalakshmi 1996). These various sources indicate that during the Early Historic period (c. 300 BC to AD 400) there were a number of centers of population and settlement growth and political and economic complexity across southern India. The Early Historic period also witnessed the spread of Buddhism, Jainism, and caste ideologies into the peninsula, with Brahmanical Hinduism appearing later, in the “medieval” seventh century AD. These ideological systems spread through a variety of mechanisms, including the expansion of northern imperial states (i.e., the Mauryan empire or the Deccani Satavahanas), economic interactions, population movements, and the founding of monastic settlements. Intensive long-distance trade, both within and beyond South Asia, and the expansion of specialized economies also characterize this period in South India. It is during this period that the South’s first urban centers appeared and state level polities emerged in multiple regions (see Champakalakshmi 1996; Thapar 1995; Rajan 1997; Allchin 1995). Processes of change in the Early Historic were neither uniform in tempo nor in geographic distribution across South India, resulting in a palimpsest of cultural variation and an archaeological and textual record that is both extremely complex and makes any large-scale generalizations about sociopolitical processes ultimately unsatisfying. A consideration of the complexities of the Early Historic period of South India is well beyond the scope of this work, as are detailed discussions of the subsequent “medieval” or “middle” period, c. AD 400–1200/1400. And, in general, these periods require far more detailed archaeological and historical study and attention to regional variability than they have yet received (see Parasher-Sen 1993). However, by the early centuries AD, we can at least put names to many of the different regions and dynasties that existed. Language communities (i.e., Tamil, Kannada, Telugu) and cultural and historical traditions that characterized different regions of the South also can be delineated; these cultural divisions would continue in importance through the Vijayanagara period (and indeed, until the present), despite considerable population movement and interaction among regions. And, as the written record expanded in subsequent centuries, it becomes relatively straightforward to trace the rise and fall and geographical extent of numerous large and small states and empires. Indeed, Stein (1998) has calculated that more than fifty dynasties are named in the South Indian inscriptional record from c. AD 400–1400.

65

66


Here, I restrict my discussion to several important polities of the centuries immediately preceding the formation of the Vijayanagara empire: specifically, the Chola, Western Chalukya of Kalyana, Hoysala, and Kakatiya states (Figure 4.1; see Nilakanta Sastri 1966; Kulke and Rothermund 1998; Stein 1998 for historical overviews of South India). These polities all belong to the late middle or medieval period, which according to conventional historical reconstructions ends in the fourteenth century AD. Vijayanagara is sometimes placed in this middle period, but has been more often seen by scholars as marking a significant transition in South Indian political, economic, and military structures and systems. Certainly, Vijayanagara’s scale and diversity were far larger than any earlier polity, and its military effectiveness and organization far greater. And at least at times, Vijayanagara’s political and administrative structure entailed greater penetration into local affairs than evident in earlier states. Indeed, Subrahmanyam (2001) has argued that South India in the Vijayanagara period and after should be viewed in the framework of broader global changes that characterized the early modern period, a development, he argues, that extends far beyond its traditional European ascription (also Talbot 2001). However we draw these divides, it is the case that to understand Vijayanagara history, we must also understand developments in the centuries that preceded its emergence.

South India from c. 900–1350 AD The four centuries preceding the emergence of the Vijayanagara empire witnessed a number of political, economic, and social changes, of which the invasions and conquests by the Delhi Sultanate were perhaps the most dramatic. From c. AD 900–1330, South India was divided under a number of competing states and empires. The central Tungabhadra River region, where the Vijayanagara capital would emerge, was claimed by the Chalukyas of Kalyani, the Yadavas (for a brief time in the early 1200s), the Kakatiyas, the Hoysalas, and the short-lived Kampili state. With the exception of Kampili, which never controlled large territories, each of these states incorporated local elites into their administrative structures in various ways, adding additional levels of complexity to this shifting political mosaic. The precise nature and extent of political control exercised by these various claimants over the region where Vijayanagara would later emerge is not at present well understood, but their very number attests to the fluidity and multiplicity of political authority in the centuries prior to Vijayanagara’s emergence.


Kakatiya Kalyani

Orugallu

Chalukyas of Kalyani

n West er

Konkan Coast

ats Gh

Kampli

Bay of Bengal

Hoysala

Kanara Coast

Pallava

Dvarsamudram Kanchipuram

Arabian Sea

Coromandel Coast

Gangaikondacholapuram

Chola Madurai

Malabar Coast

0

km

Pandya

200

4.1 Core areas of pre-Vijayanagara states of South India (after Stein 1998)

As Vijayanagara rule was extended across South India after c. AD 1350, the rapidly expanding empire perforce incorporated many political and economic practices of these earlier states. It is thus not surprising that scholars seeking the antecedents of various Vijayanagara structures and practices have suggested many different sources. Historians such as Burton Stein (1980), Karashima (1992), and Nilikanta Sastri (1966) have emphasized

67

68


Vijayanagara’s links to the Tamil Chola empire. In contrast, Cynthia Talbot (1994, 2001) and later writings by Stein (e.g., 1989a) have suggested that we look to the inland states of the southern Deccan, and particularly the Kakatiyas and Hoysalas, respectively, for the sources of many characteristic Vijayanagara traits. More recently, Talbot, Richard Eaton (2000), and Philip Wagoner (1996, 2000) have also argued that Vijayanagara scholars must also consider the influences of and Vijayanagara’s connections to the Islamic polities of the period – including the Delhi Sultanate and later Bahmani State, and its successor Deccani sultanates. Similarly, in discussions of the origins of the founding dynasty of the empire, multiple claims have also been made – linking the Sangamas to the Kampili, Kakatiya, or Hoysala states (Subrahmanyam 1998c; B. Stein 1989a). Given the geographic extent and regional and cultural diversity of the Vijayanagara empire, the flexibility of Vijayanagara rule, and the global stage in which Vijayanagara participated (see Sinopoli 2000), it seems certainly the case that all of these polities and regions contributed to developments at Vijayanagara in various ways and in various times and places. Further, the fact that the precedents of Vijayanagara political structures and practices are so difficult to trace may in itself be evidence of the empire’s success in laying claim to universal status. Below I provide a brief outline of the historical setting of major pre-Vijayanagara states, and their connections to later South Indian developments.

The Cholas Of the pre-Vijayanagara South Indian polities, the Chola empire (c. AD 850–1279) has been the most extensively studied by historians (see for example, Champakalakshmi 1979, 1995, 1996; Dehejia 1990; Hall 1980, 1994; Heitzman 1995, 1997; Karashima 1978, 1984; Nilakanta Sastri 1955; Stein 1980). The Chola heartland lay in the rich Kaveri delta region of southern Tamil Nadu (Figure 4.1), an area of high agricultural potential, particularly for wet rice agriculture, which could support high population densities. The Chola dynasty first appears in historical sources in the Çangam literature (c. 200 BC–AD 200) as one of three major “peoples” or lineages of the Tamil region (along with the Chera and Pandya). A dynasty of the same name re-emerged in the Kaveri delta region in the ninth century AD, although with no clear relation to the ancient Cholas (Heitzman 1997: 3). Chola territories reached their maximal extent during the reign of Rajaraja I (AD 985–1015), whose military victories expanded the empire to the west,


south, and north. The Cholas remained the largest polity in the region until the late twelfth century, when they lost territories and political authority as a result of the expansion of the Pandya state to their south, uprisings within Chola territories, and conflicts at the imperial center. While both significantly earlier and geographically removed from the Vijayanagara core area, the Chola polity was the largest of the pre-Vijayanagara South Indian states and has figured prominently as a comparandum for Vijayanagara in South Indian historiography. And it is certainly the case that the affluent and fertile Tamil-speaking areas that the Cholas had dominated were an important focus of Vijayanagara imperial expansion and economic intensification. The upland peripheries of the Chola heartland, in northern Tamil Nadu and southern Andhra Pradesh, also became a focus of urban and industrial investment and expansion throughout the Vijayanagara period, and especially in the sixteenth century, particularly for textile production. It is from these areas, where a great deal of epigraphic research has focused, that some of our most valuable inscriptional evidence for changing roles of craft producers is found (see Karashima 1992; Ramaswamy 1979; chapters 6 and 7, this text). The Chola period also provides evidence for several features that would come to characterize later centuries. These include trends toward urbanization, monetized economies, and economic intensification in both agricultural and industrial (particularly textile) production. The Chola period also witnessed the beginnings of the complex interactions between temple centers and regional and imperial nobility that would become so prominent in the Tamil region and beyond during the Vijayanagara period (see discussion later this chapter). These interactions involved substantial donations made by a variety of individuals and communities, temple investment in production and agrarian expansion, and the role of temples in political legitimation (see Appadurai 1981). Like later Vijayanagara, and indeed as in all historic empires, multiple competing sources of power existed in the Chola polity and these changed over time and varied over space in their influence and effectiveness. In a recent study based on Chola inscriptions, Heitzman (1997) has traced the temporal development of the Chola state and the diverse sources of power within it – including the ruling dynasty, regional elites, temples, and agrarian laborers, artisans, and merchant guilds. Heitzman has argued for regional and temporal variability in Chola state structure, with evidence for increased penetration of state administration into local structures following the consolidation of state power during the later Chola period.

69

70


The Deccani states The Deccani states, the Chalukyas of Kalyani, Hoysalas, and Kakatiyas, have received far less systematic study than the Cholas. It is, however, increasingly evident that in the Deccan as well, the tenth to thirteenth centuries were marked by increasing political and social complexity, as well as by agrarian and demographic expansion into dry upland regions (Stein 1989a: 21), processes that intensified in subsequent centuries and have often been viewed as Vijayanagara innovations. Indeed, writing about the Kakatiyas, Cynthia Talbot (2001: 173) has observed: Many of the developments of the late Kakatiya period – its growing militarism, the disappearance of ancient lineages of chiefs, the migration of warriors into areas outside their places of origin, the emphasis on personal ties between lord and subordinate, nayankaramu revenue assignments, and the greatest control over distant localities exercised by kings through their delegated officers – are specified in the secondary literature as innovations of the Vijayanagara period.

This observation can also be extended to other Deccani states.

The Chalukyas of Kalyani The Chalukyas of Kalyani first appear in the inscriptional records as subsidiaries of the Rashtrakuta state, which ruled much of Karnataka from the eighth through the tenth centuries AD. The Chalukyas of Kalyani claimed descent from the earlier Chalukyas of Badami (sixth through eighth centuries AD), though the precise nature of this relation is not known. In AD 973, led by Taila II, the Chalukyas overthrew the Rashtrakutas and became rulers of an independent state. Throughout their two-century reign, Chalukya forces battled the Cholas, Hoysalas, Kakatiyas, Yadavas, as well as several lesser dynasties that vied for hegemony across the south. Along with numerous temples (Patil 1992), the Chalukyas left behind thousands of inscriptions and several Kannada and Sanskrit literary works that provide insights into their geographic and temporal extent and political ideology. Rulers claimed imperial status and adopted as their emblem Varaha, the boar incarnation of Visnu, an emblem later adopted by the Kakatiya and Vijayanagara dynasties (Figure 4.2). Chalukya administrative structures are hinted at in inscriptions that include the title “head of the 72 departments.” Mention of many of these “departments” also appears in a contemporary Kannada political treatise, including departments for villages, justice, stores, treasury, forts, robes, ornaments, wages, jails,


4.2 Vijayanagara boar emblem from gate at Chandragiri

frontiers, and doorkeepers, among others (Basavaraja 1983: 68–70). Exactly how these units were structured, what their responsibilities were, and how far they penetrated into state territories and local administrative traditions has not been determined. The Chalukya regime began to disintegrate in the mid-twelfth century, as the Yadavas emerged as a power to the north, the Hoysalas to the south, and the Kakatiyas to their east. Each of these formerly tributary dynasties became a center of an independent state that ruled competing territories in the late twelfth through thirteenth centuries. Here I consider only the Kakatiyas, Hoysalas, and Kampili states, which ruled territories that later came to be dominated by Vijayanagara.

The Kakatiyas The Kakatiya lineage of Andhra Pradesh first appears in the inscriptional record in AD 956 as a tributary of the Rashtrakutas and later of the Chalukyas of Kalyani (Parabrahma Sastry 1978: 2; Raghunadha Rao 1993). In c. AD 1160, under the ruler Rudradeva, the Kakatiyas declared independence from the weakened Chalukya state and became rulers of an independent polity. Like Vijayanagara would two centuries later, the Kakatiyas began a period of expansion from their inland base. And, with varying success, they ruled for nearly two centuries over much of Andhra Pradesh and parts of Karnataka and northern Tamil Nadu from their capital at Warangal (Orugallu). While little archaeological research other than temple documentation (Wagoner 1986) has been conducted on this period, a large number of inscriptions

71

72


(c. 1500 from Andhra Pradesh; Talbot 2001: 25) and some Telugu literary works have been studied. Roughly 200 inscriptions of the period explicitly acknowledge Kakatiya rule; many more do not (Talbot 1994, 2001). Together, these sources provide ample evidence for agrarian expansion, the increasing importance of (Shaivite) temples as recipients of donations and investors in agriculture and craft, the importance of maritime trade, and the emergence of new categories of Telugu administrator-warriors. Kakatiya rulers, administrative and military elites, temples, and local communities were all actively involved in the expansion of agriculture in the state’s upland core, sponsoring the construction of numerous irrigation reservoirs and canals and the founding of agricultural settlements (Parabhama Sastry 1978: 198–202). Reservoir irrigation was particularly important in creating pockets of relatively densely settled fertile agricultural lands in the semiarid uplands of the Kakatiya core area. Mobile pastoralist communities were also widespread in these regions, and played an important role in the transportation of agricultural and other commodities (Talbot 2001: 24). Accompanying this economic expansion were increases in revenue collection and taxation. Kakatiya inscriptions list a range of taxes and discrete occupational groups of the period, providing evidence for a complex and diversified economy, in which maritime trade and specialized craft production played important roles. Along with taxes on land and agricultural produce, taxes on trade in textiles and other goods provided important revenues to the state. Both maritime and overland trade is evidenced. Much of this trade was under the control of merchant organizations, such as the “Ayyavole 500” (see pp. 103–105), which were based in emerging market and temple towns. There is some evidence that members of occupational groups such as potters, weavers, and oil merchants also were organized in corporate groups. Inscriptions of the Kakatiya period also provide evidence for a number of levels of administrative and revenue-collecting units, from village and occupational groups and other corporate bodies, to territorial units and royal officers. In a comprehensive study of all of the recorded inscriptions in which the Kakatiya dynasty is explicitly acknowledged, Talbot (1994) has documented two categories of royal subordinates. The first are nobles whose claim to authority comes from their royal genealogies and hereditary leadership in areas incorporated within the Kakatiya state. Talbot suggests that these locally powerful nobles were very important to Kakatiya state-building and expansion in the first century of their rule, and were incorporated into the state through conquest and alliance (Talbot 2001). The second category of royal subordinates documented by Talbot consists of individuals she calls “officers.” These officers lacked hereditary claims to


power, and instead derived their authority from personal relationships with the ruler. Many were warriors. The officers became increasingly prominent in the Kakitiya inscriptional record after AD 1261, beginning during the reign of Queen Rudramadevi. Talbot sees this as evidence for attempts by the last two Kakatiya rulers to create a new political infrastructure that undercut the competitive claims and authority of hereditary regional elites. Approximately twenty-five titles are associated with the officers, including minister, accountant, tax collector, army commander, etc. The most common title in the inscriptions is “nayaka,” or nayankapuvarum (“holders of nayankara land grants”), which first appears in the Telugu region during the Kakatiya period. Nayankara were land grants ceded to individuals, mostly warriors, by the Kakatiya ruler as a reward for their service to the court. Within the nayankara territories, nayakas could collect, remit, and dispose of tax revenues. Beyond this, nothing is known of the nature of the nayakas’ financial or military obligations to the state, nor of the duration of their positions. Whatever the precise nature of the connection between the Kakatiya nayankara holders and the reappearance of this office in the subsequent Vijayanagara period, at the very least we see striking evidence of “the growing impact of Telugu warrior society” (Talbot 1994: 262) across South India, a trend that would intensify dramatically during subsequent centuries.

The Hoysalas Like the Kakatiyas to their east, the Hoysala dynasty first appears in the inscriptional record as subordinates of the Chalukyas. Although their origins are obscure, they are believed to have been the ruling family of a small Kannada polity that became allied with the Chalukya ruling dynasty through marriage and political alliance at the start of the eleventh century (Muddachari 1972: 59). By the second half of the eleventh century, several inscriptions indicate that the Hoysalas were becoming powerful and influential in their own right. Throughout the twelfth century, successful Hoysala rulers continued to acknowledge Chalukya sovereignty even as they increased their autonomy and territorial extent. The Hoysalas expanded into Tamil regions and engaged in complex relations of political diplomacy and warfare with the Pandyas and Cholas. By the mid-twelfth century the Hoysala polity was an imperial state ruling over extensive territories; their imperial claims are evident in titles such as “The Emperor of Brilliance” or “Overlord of Great Kings” (Derrett 1957: 209–211) that appear in the inscriptional record at this time. By AD 1200, acknowledgments of the Chalukyas disappeared entirely (Muddachuri 1972: 61).

73

74


The Hoysala state remained a significant southern power until the sultanate invasions in AD 1310. At their maximal extent in the late twelfth century, the Hoysalas ruled over territories that included all of modern Karnataka and parts of Tamil Nadu as far south as the Kaveri River basin. As Vijayanagara would later be, the Hoysala capital Dvarsamudram (Halebid, Hassan District) was located in rugged, easily defensible, uplands at the northern boundaries of their state. While the elaborate Hoysala temple styles and Kannada literature have been studied in considerable detail, there has been relatively little archaeological or historical research focusing on Hoysala political or social organization. Titles in the inscriptional record suggest a complex social and political hierarchy with a variety of administrative and military offices. Several ministers are known from inscriptions, with a variety of titles. The title dandanayaka was used for high-status military officers and regional governors, while nayaka appears to refer to lower status military leaders (Derrett 1957: 188). To date there has been little scholarship focusing on the precise relations among offices and administrative units. Agricultural expansion, particularly through the construction of irrigation reservoirs, is documented in many Hoysala inscriptions. Temple donations by a variety of communities suggest the increasing prominence of those institutions. Merchant organizations appear as donors in several twelfthand thirteenth-century inscriptions (Rangaswamaiah 1972); as discussed in this chapter, these influential organizations were engaged in trade of a range of goods within South India and in international maritime commerce.

Kampili Kampili is the shortest-lived and smallest of the pre-Vijayanagara polities discussed here. Its importance lies in its location, very near to where the Vijayanagara capital would later be founded, its relations with the Delhi Sultanate, and in suggested historical connections between Kampili and Vijayanagara. In particular, several scholars have suggested that the reputed founders of Vijayanagara, the Sangama brothers, had been officers in the Kampili military. The rulers of Kampili began their careers as tributaries of the Yadava kings of Devagiri, the first Deccani polity defeated by the Delhi Sultanate. Following Sultanate occupation of Devagiri, the Yadava general Muhammad Singa fled with his son Kampili to the northern shores of the Tungabhadra River, where he established a base from which to resist Sultanate forces (Stein 1989a: 18). Kampili was a powerful military leader in his own right, and


became the founder of this small war state. Accounts of Kampili-Raya (King Kampili) and his military prowess are known from several contemporary Arab historians as well as from numerous later Kannada texts (Narayana Rao, Shulman and Subrahmanyam 2001: 101–115). Not surprisingly, these texts present rather different perspectives, but all acknowledge the successes and eventual brave defeat of Kampili and his forces. Several inscriptions also document the history of Muhammad Singa, Kampili-Raya, and Kumara Ramanatha, Kampili’s son (Patil 1991a). For approximately fifteen years, the forces of Kampili-Raya successfully resisted Sultanate attacks. In 1327, after two failed attempts, the Sultanate army killed Kampili-Raya, and his kingdom collapsed. According to the texts, at this third siege it soon became clear that defeat was inevitable against the much larger and better-armed Sultanate army (one Kannada text estimates the size of this army in the hundreds of thousands). Rather than be captured, the noble women inside the king’s fort committed suicide, and Kampili and his forces ventured out to fight to their deaths. The precise location of Kampili’s forts has been a matter of some controversy. Several sites are mentioned in the texts: Haneya, Dorvadi, Kummata, and Hosamaledurga (Patil 1991a: 179). Of these, Kummata and Hosamaledurga appear to have been the most important centers for Kampili-Raya’s activities. The latter has often been associated with Anegundi on the northern bank of the Tungabhadra River, though Patil (1991a) has questioned this interpretation and places it approximately twenty kilometers to the south. Kummata has been recently identified, in the rugged hills some ten kilometers to the north of the Vijayanagara urban core (Patil 1991b). Although Kampili’s resistance to the Delhi Sultanate became a powerful political symbol in the region, we know little about the political or social structures of his polity. In any case, it is probably best to view KampiliRaya’s major accomplishment as having been an effective military strategist, who outlasted the Yadavas, Hoysalas, and Kakatiyas in resisting the Delhi Sultanate. I turn to the latter below.

The Delhi Sultanate In the late thirteenth century, the polities and populations of South India long used to centuries of warfare and alliance building, faced a powerful new enemy from the north: the military forces of the Sultanate of Delhi (see Nilakanta Sastri 1966: chapter 11; Michell and Zebrowski 1999: 4–7; Jackson 1999). The general Alauddin led the first Sultanate foray into peninsular India in AD 1296, against the Yadava capital of Devagiri (Daulatabad,

75

76


Maharashtra). After a brief siege, the Yadavas capitulated and paid a large ransom to the Sultanate. Alauddin and his forces returned to Delhi, where the general was soon declared emperor. Alauddin’s second, unsuccessful, expedition to the South took place in AD 1302–1303, and was directed against the Kakatiya capital Warangal. Following a resounding defeat, it was six years before sultanate forces returned. This time, Warangal was defeated and became tributary to the Sultanate. In AD 1310, the Hoysala forces fell to the Sultanate’s armies, which then proceeded further south to conquer Pandya territories. Thus, within fifteen years of the Sultanate’s initial foray into the peninsula, the Yadava, Kakatiya, and Hoysala states had all been forced to acknowledge its suzerainty. The Sultanate armies did not initially occupy the territories they conquered and were instead content with the flow of tribute to the Delhi capital. However, following the accession of a more resistant Yadava king in AD 1312, Delhi forces briefly occupied Devagiri, which they quickly lost and then reconquered in AD 1317. Areas of resistance to Sultanate rule persisted throughout the south, particularly in the Tungabhadra region where first the Kampili state emerged, and later, Vijayanagara. In AD 1321 the Tughluq family claimed the Sultanate throne and, taking advantage of the resultant disorder in Delhi, many of the southern regions stopped forwarding tribute payments to Delhi. In AD 1325 the second Tughluq ruler, Muhammad Shah, took the throne and sought to reconsolidate control over the Deccan and Tamil regions. In an attempt to impose a greater presence in the South, the capital was shifted from Delhi to Daulatabad (the former Devagiri), an unsuccessful effort that lasted only a few years. Sultanate authority in the south, never strong, further weakened in the next several years, as the Sultanate itself began to fragment. However, the earlier southern states were by then dangerously weakened or destroyed and were unable to reconsolidate their authority: the Kakatiya state collapsed in AD 1323; the Hoysalas fell shortly thereafter; and the small Kampili polity fell in 1329. In the wake of the Sultanate’s withdrawal at the end of the 1320s, the political landscape of South India had been dramatically reshaped. None of the region’s major earlier polities survived. To the north of the Tungabhadra River emerged the powerful Bahmani Sultanate (AD 1347–1526), which by the early 1500s had fractured into five smaller polities – the Deccani sultanates of Golconda, Berar, Bijapur, Bidar, and Ahmadnagar (Figure 4.3). And to the south emerged Vijayanagara – which began as a small polity based on the southern back of the Tungabhadra River less than a dozen kilometers south of Kampili-Raya’s Kummata fortress.


AHMADNAGAR

BIRAR

Ahmadnagar

GOLCONDA BIJAPUR Bijapur

ST E WE

Konkan Coast

Goa

Golconda

VIJAYANAGARA Tadpatri

RN

Honawar

BIDAR

Bidar

Penukonda Tirupati

Bhatkal Basrur

G HA

TS

Kanara Coast

Bay of Bengal

Tiruvannamalai

Coromandel Coast

Calicut

Arabian Sea Madurai Cochin

Malabar Coast

0

km 200

4.3 The Deccani sultanates and Vijayanagara empire

It is difficult to overestimate the long-term consequences of the Delhi Sultanate’s incursions on political developments in Southern India. Nonetheless, the constant warfare that characterized their thirty-year presence in the region, as well as their emphasis on tribute over institutional engagement, meant that they had relatively little impact on regional political or administrative structures. The major exception was in their military

77

78


practices, particularly the reliance on cavalry, which was adopted by the first Vijayanagara rulers and became a key component of their military effectiveness. But the Delhi Sultanate did not impose significant new administrative systems, ideological structures (though certainly some conversion to Islam occurred across their territories), or economic practices on the polities they conquered. Instead, they relied on military power and the application or threat of force to extract revenues from the region.

V I JAYA NAG A R A As noted, the small state of Vijayanagara emerged in the wake of the withdrawal of the Delhi Sultanate, and the collapse of the earlier Deccani states. It expanded quickly, and within a few decades had become South India’s largest historic empire. In the remainder of this chapter, I provide an overview of Vijayanagara history, geography, religion, and social life, along with a summary of various theoretical perspectives on its organization.

Dynastic history Four dynasties ruled the Vijayanagara empire: the Sangama (c. AD 1350– 1486), Saluva (c. AD 1486–1505), Tuluva (c. AD 1505–1569), and Aravidu (c. 1569–1654) dynasties.2 The first three were based at the eponymous capital of Vijayanagara. The Aravidu kings ruled a shrinking empire from three successive capitals following Vijayanagara’s abandonment in AD 1565. Several contemporary stories recount the history of the empire’s beginnings (Subrahmanyam 1998c; also Sinopoli 2000). They differ on the origins, ethnic background, and prior political affiliation of the five Sangama brothers, the empire’s reputed founders. While the truth of Sangama origins is presently unresolvable, several things are clear. The first Sangama rulers, brothers Bukka I (c. AD 1352–1377) and Harihara II (c. AD 1377–1404), founded their state in a relatively sparsely populated area along the Tungabhadra River. There they built their capital, Vijayanagara, the “City of Victory.” From this base, the Sangamas quickly began a period of outward expansion and internal consolidation. Their expansion was made possible by effective military forces that relied on technologies introduced by the earlier invading armies of the Sultanate, particularly cavalry. Their state expanded rapidly to the south and by the late fourteenth century had reached imperial stature. As the empire grew, its capital also expanded rapidly; by the


beginning of the fifteenth century, it extended over 15-square kilometers, with as many as 100,000 inhabitants (see chapter 5, pp. 142–145). Like the South Indian states before them, early processes of Vijayanagara imperial consolidation responded to a variety of local conditions and involved a combination of collaboration with local elites and the imposition of Sangama kin as regional rulers. The early Sangamas endowed temples throughout their expanding territories and encouraged the expansion of reservoir-based agriculture in the semiarid upland regions around their capital. While in the early decades of their rule the empire was loosely structured into semi-autonomous regions governed by Sangama kin, over time the Sangamas began to institute a more formalized system of regional officers (Stein 1989a). They were successful in collecting significant revenues through tribute and taxation. Nonetheless, the ability of the center to successfully dominate more distant regions remained relatively limited, and was closely correlated to the perceived strength and military effectiveness at the empire’s center. The Sangama period was marked by nearly constant warfare with polities to both the south and the north. To their north and northeast, Vijayanagara forces engaged in repeated battles with the armies of the Bahmani Sultanate and the Orissan Gajapatis. Despite occasional victories, Vijayanagara’s attempts to expand northward were not successful for long – a pattern that was to persist throughout the empire’s history. Their southward expansion was much more effective. By the late 1300s, Vijayanagara had claimed sovereignty over all of Karnataka and Andhra Pradesh, and much of the Tamil South and portions of the Arabian sea coast on the west. The southwestern Kerala coast lay outside of their direct control, as it would throughout the empire’s history (B. Stein 1989a; Nilakanta Sastri 1966; Morrison 1995). Despite this rapid expansion, many of the early military successes of the first Sangama rulers were short-lived, and later kings were repeatedly forced to retake territories that had been lost by various of their predecessors. In the early fifteenth century, two highly effective Sangama rulers, Devaraya I (AD 1406–1422) and Devaraya II (AD 1424–1446), succeeded in retaking and solidifying control over these extensive territories. They oversaw a period of political consolidation, based primarily on an expanded and highly effective military (B. Stein 1989a), including the incorporation of a large number of Muslim mercenaries. However, after the death of Devaraya II, the Sangamas and the empire they ruled entered a period of dramatic decline, as territories and imperial resources were lost to internal uprisings and external threats. That the late fifteenth century was a time of crisis in the empire and throughout South India is amply evidenced in changes in the inscriptional

79

80


record. Talbot (2001, n.d.a, n.d.b), Karashima (1992), and Morrison and Lycett (1994, 1997; also Morrison 1995) have all documented dramatic decreases in the number of donatory inscriptions at this time in Andhra Pradesh, Tamil Nadu, and central Karnataka, respectively. This broad pattern across imperial territories reflects both a general decline in state authority and an overall drop in the wealth available for economic investment across the south. As discussed later this chapter, donations to temples both require wealth and are closely associated with investment in the expansion of agriculture and craft production and commerce. Their dramatic decline in this period of political uncertainty suggests that there was not simply a replacement of imperial elites by local elites as the power at the center weakened. Instead, imperial weakness appears to have contributed to a far more pervasive period of instability and economic decline that affected much of the South. This decline is marked also in archaeological evidence at the capital; the Vijayanagara Metropolitan Survey project has identified a significant decrease in construction activities in the region throughout the fifteenth century. Responding to weakness at the center during the last decades of Sangama rule, the military commander Narasimha Saluva, in charge of the fortress town of Chandragiri to the southeast of the capital, began to expand his powers and act autonomously of the Vijayanagara court. At the same time, threats came from the North, as the armies of the Orissan Gajapati state attacked the Vijayanagara capital in AD 1485. The Gajapatis succeeded in conquering and, for a brief period holding, much of India’s southeastern coast, from Orissa to the northern shores of the Kaveri River. On the opposite coast, the Bahmani Sultanate gained control of many of Vijayanagara’s Arabian Sea ports. In 1486, Narasimha Saluva took control of the Vijayanagara army and declared himself ruler, founding the empire’s second, Saluva, dynasty. During his seven-year reign, he retook many of the territories that had been lost under the last Sangama rulers. His success was short-lived as he and his successor, Narsimha II, faced repeated attacks by the Adil Shahis of Bijapur, who had emerged as a major power to the north in the wake of the collapse of the Bahmani Sultanate. In 1501, the Saluvas disappeared as they had emerged, following an internal takeover by a regional military commander, Narasa (Narasimha) Nayaka. He became the first ruler of Vijayanagara’s third dynasty: the Tuluvas. The Tuluvas ruled from the city of Vijayanagara until AD 1565. During the first forty years of their reign, powerful Tuluva rulers oversaw the reconsolidation and expansion of the empire and capital (Figure 4.3). Under the


rule of Krishnadevaraya (AD 1505–1529), the empire’s most effective ruler, a period of administrative restructuring occurred. Krishnadevaraya’s innovations resulted in the development of more effective systems of rule, including, in some areas, an increased reliance on Telugu military leaders or nayakas and Brahman administrators who displaced existing indigenous elites. The Tuluva period was a time of dramatic expansion both demographically and economically, including in agricultural and industrial production, commerce, and the flow of resources into imperial coffers. These changes are evidenced in the enormous increase in inscriptions recording temple donations, which reached their highest frequency in first half of the sixteenth century (Talbot 2001; Morrison 1995; Morrison and Lycett 1997; Karashima 1992). The Tuluva reign was also a time of urban expansion at the capital and throughout the empire, with large-scale construction of temples, irrigation works, fortifications, and other facilities. The capital reached its maximum size in the early sixteenth century. The city core extended to more than 30square kilometers and the population in the city and hinterland likely exceeded a quarter million inhabitants. The Vijayanagara of Krishnadevaraya would become a city of legend to later South Indian poets and bards, famous as the seat of political and sacred authority (see Mack 2000, 2002; Sinopoli 2003) and as a center of wealth and luxury. During the reign of Krishnadevaraya’s successor, Achyutadevaraya (AD 1530–1542), weaknesses in the empire began to resurface. As in earlier periods, internal weakness and external threats coalesced to challenge imperial sovereignty. Disputes arose at the center over Achyutadevaraya’s succession at the same time that resistance to imperial authority spread throughout the empire, led by some of the powerful nayakas who had been placed in office (or acknowledged) by earlier imperial acts. From outside the imperial borders came challenges from the five Deccani sultanates that had emerged in the wake of the collapse of the Bahmani state. A succession struggle followed the death of Achyutadevaraya. At its end, Rama Raya, son-in-law of Krishnadevaraya, ruled as regent for the child-king Sadashiva (1543–1569). Rama Raya and his brother Tirumala did not relinquish power as Sadashiva grew to manhood, and they led the Vijayanagara armies in some ill-planned attacks against their northern foes (B. Stein 1989a). In AD 1565, the sultanates of Bijapur, Golconda, and Ahmadnagar united against the Vijayanagara military and defeated them in a major battle that took place some 100 kilometers north of the imperial capital. Rama Raya was killed and Tirumala appointed himself regent. The court shifted south to Penukonda, abandoning Vijayanagara, which was never substantially reoccupied.

81

82


In AD 1569, Sadashiva was murdered and Tirumala was crowned king, transferring rule to the empire’s fourth and final dynasty: the Aravidus. Over the next 100 years, Aravidu kings ruled over shrinking territories from the capitals of Penukonda, Chandragiri, and Vellore. Throughout this period, a number of successor states emerged in Andhra and Tamil Nadu. These “nayaka” states were ruled by military elites who had risen to power during the Saluva and Tuluva reforms. While the rulers of these states initially proclaimed their allegiance to the Vijayanagara throne, throughout the late sixteenth and early seventeenth centuries they became largely autonomous, establishing their own bases of power as Vijayanagara’s authority and legitimacy disintegrated. The last Aravidu king died in AD 1654, bringing an end to even the name of Vijayanagara.

Imperial geography As evident in the previous discussion, territories shifted in and out of imperial boundaries with considerable rapidity during the approximately 300 years of Vijayanagara hegemony. Many areas claimed by the empire did not necessarily acknowledge those claims, or had at best marginal links to the imperial center. It becomes very difficult therefore to parse out precise imperial boundaries, much less the boundaries of various administrative units, although it is clear from a variety of sources that these were numerous. In this section, I present a general overview of the geographic and environmental settings of the regions where Vijayanagara history played out. These include the areas of the modern states of Karnataka, Andhra Pradesh, and Tamil Nadu. Coastal Kerala, to the southwest, never fell under direct Vijayanagara authority and remained a region of relatively small-scale autonomous polities and tribal peoples throughout the period. Stein (1989a: 2) has estimated that at its peak the empire encompassed an area of approximately 360,000square kilometers, and a population of some 25 million. This enormous area varied tremendously, culturally, historically, and environmentally. Like the Kakatiyas, Hoysalas, and Western Chalukyas, the Vijayanagara state emerged in South India’s inland semiarid uplands, a harsh area with unpredictable rainfall that poses considerable challenges to agrarian expansion. The Sangamas quickly focused their expansion efforts to the southeast, to the fertile rice-growing river valleys and deltas of Tamil Nadu and to the seacoasts on the east and west, where much of South India’s wealth was concentrated. As in many historic empires (see Alcock et al. 2001; Sinopoli 2001a), the territories claimed by Vijayanagara were discontinuous and


formed a complex mosaic that changed over time as population movement and agrarian expansion transformed forested areas inhabited by mobile “tribal peoples” into zones of agriculture and permanent settlement. Below, I briefly characterize these different regions.

The upland core Vijayanagara expansion began from the Sangama brothers’ newly founded capital on the Tungabhadra River in the semiarid uplands of central Karnataka (see chapter 5 for a more detailed description of Vijayanagara and its immediate hinterland). Around this locale, the core of the empire took shape – a zone that Stein (1989a: 58) has estimated as ultimately covering about 30,000 square kilometers, extending from Karnataka’s Bellary District, where the capital is located, east to Kurnool and Cuddapah districts of Andhra Pradesh (Figure 4.4). This extensive area was the most secure and durable focus of Vijayanagara authority throughout the first two hundred years of the empire’s history. Geologically, this area comprises the southern extension of the Deccan plateau and is dominated by zones of rugged granitic hills interspersed by productive pockets of arable soils (Figure 4.5). In the western part of this region, the terrain is more rugged; the plateaus vary in elevation from c. 450 to 600 meters in the Tungabhadra basin, with surrounding ranges rising an additional 300 meters. Further east, the terrain becomes more open, with more extensive areas of arable soils. With irrigation, much of this area, particularly the eastern zones dominated by clay-rich “black cotton soils,” could be highly productive. There are no navigable rivers in the core region of the empire. The Tungabhadra River, along which the capital lay, is a tributary of the Krishna. These rivers originate in the Western Ghats and flow east into the Bay of Bengal. Like other smaller rivers of the uplands, the Tungabhadra is monsoon-fed and seasonally dry. Because of the region’s rugged topography, the floodplain of the Tungabhadra is in many areas extremely narrow; canal irrigation is limited to those areas with narrow bands of low-lying soils. An important agricultural zone near the capital was the region north of the Tungabhadra River extending to the Krishna River. This area, the Raichur Doab (meaning land between two rivers), was a zone of contention throughout the Vijayanagara period and control over it changed frequently, and violently. The inland core region averages between forty and seventy-five centimeters of rainfall per year; the vast majority falls during the summer monsoon. Rainfall is irregular and unpredictable. Famine periods were not

83

84


Vijayanagara

3000 2000 1000 500 200 0 0

km

300

meters asl

4.4 The Vijayanagara empire: upland core

uncommon; documented famines of the Vijayanagara period occurred in 1396, the early 1420s, 1471–1472, the 1540s, and the early 1630s (Morrison 2000: 30–31; Subrahmanyam 1990a: 359). Because of the region’s aridity and rugged topography, agricultural production in this area relied primarily on reservoir irrigation and seasonal rainfall. Kakatiya and Hoysala investments in reservoir construction in their core areas were noted earlier; such constructions continued in the uplands throughout much of the Vijayanagara period, dramatically increasing agricultural productivity in the empire’s upland core (see Morrison 1995, in prep a, in prep b). Along with increased subsistence production near the capital, the inland districts of Andhra Pradesh became among the subcontinent’s major cotton-producing regions, and it is


4.5 The Vijayanagara region: the Tungabhadra River valley

here that much of the region’s textile production was focused (see chapter 6; Ramaswamy 1985c). This region was also rich in iron ore sources. To the south of the capital, the Sandur hills are productive sources of hematite and iron ores that are still mined today. Other major ore areas were located in western Andhra Pradesh and in western Karnataka.

The river valleys of the southeast Southeastern India (Tamil Nadu and eastern Andhra Pradesh) is a dissected environment, dominated by several major river systems that flow into the Bay of Bengal. Coastal port towns were major trading centers, involved in commerce with settlements along the Bay of Bengal, as well as in East and Southeast Asia, and Arabian seaports (Figure 4.6). This region is bordered on the inland side by the Eastern Ghats. Unlike the more dramatic Western Ghats along the Arabian Sea coast, the Eastern Ghats are low and discontinuous, rising only a few hundred meters above sea level. Topography of the region slopes from semiarid uplands in the west down to the extensive coastal plain in the east. Several large rivers traverse the region, defining

85

86


Golconda Bijapur

Kris

Goa

hna

VIJAYANAGARA

Honavar

n Tu

Bhatkal Barkur Basrur

ra ad bh a g

Pe n nai

Penukonda

Chitradurga

Ikkeri

Tirupati Chandragiri Vellore

Mangalore

Pulicat Madras Senji

Srirangapatnam

Puducheri

Kaveri

Cananore

Srirangam

Calicut

Thanjavur

Cochin

Madurai

Kollam

Tuticorin Tirunelveli

Major overland routes

O

km

300

4.6 Vijayanagara-period seaports and major overland routes

Nagapattinam


major routes of communication and creating fertile zones of broad east–west alluvial valleys. The flood plain is broad, as much as 80 to 100 kilometers across in the fertile Kaveri River Delta. Extensive mangrove swamps lie along the shorelines and deltas. Rainfall in this zone ranges from about 60 to 100 centimeters per year, with lower precipitation rates occurring in the higher areas and to the north of the Kaveri River (Mani 1974: 137). This region benefits from both summer (southwest) and winter (northeast) monsoon rains. The lowland river valleys and delta of Tamil Nadu supported dense populations under the Cholas and Pandyas, several centuries before the Vijayanagara’s emergence. Rice cultivation predominated, facilitated by canal irrigation and extensive paddy construction. By the early 1300s, dense agricultural settlement and multiple urban centers and temple towns characterized this region (Ludden 1985), a pattern that intensified throughout the Vijayanagara period. The upland areas of the southeast were more sparsely populated. Before AD 1300, there were small pockets of rice cultivation centered on upland reservoirs; however, much of the region consisted of uncultivated semiarid forestlands occupied by small and sparsely distributed tribal communities and pastoralists. The Vijayanagara period witnessed major changes in these historically marginal upland areas. They became important foci for agricultural and demographic growth, fueled by the expansion of Teluguspeaking communities from the north. Forested regions were transformed into zones of agricultural production through the construction of irrigation facilities and the resettlement of agricultural communities. Such facilities were sponsored by temples, regional elites, and by the late fifteenth century, Telugu-speaking warriors and nayakas. The majority of these irrigation facilities were reservoirs. Agricultural production increased in the lowland regions as well, through the expansion of canal systems. The upland areas of the southeast were a focus of cotton production and the production of important dye plants such as indigo. Silk was an important product of the lower lying areas. As discussed in chapter 6, the urban centers and temple towns of this region were an important focus of South Indian textile manufacture.

The Western Ghats The west coast of the southern Indian peninsula differs dramatically from the east. Here there is no broad coastal plain fed by fertile rivers. Instead, a narrow coastal plain is abutted by the imposing Western Ghats, which rise as much as 1,200 meters over a horizontal distance of only two to three

87

88


kilometers. Upland plateau areas to the east are accessible only through a small number of narrow alluvial valleys that cut through the ghats. As a result, the west coast has long been largely isolated from developments to its east, and, as noted earlier, this region fell outside of direct Vijayanagara control. The Western Ghats also serve as a barrier to rainfall. More than 250 centimeters of rain fall per year on the western (ocean) side of the ghats, nearly five times the average rainfall than on the upland plateaus to their east. The western slopes of the ghats support a dense tropical forest with diverse vegetation that varies somewhat from north to south. Botanists working in the region have documented more than 3,500 species of flowering plants. Most important for human consumption are valued woods such as sandalwood and a variety of spices. Pepper is indigenous to the southern coastal region, and this area was the peninsula’s major source of this highly valued crop during the Vijayanagara period. Other important products included cardamom, ginger, honey, wax, and a wide range of medicinal and aromatic plants (Morrison 2001). Historically, upland tribal communities procured these products and engaged in complex economic relations with urban dwellers (Morrison 2001). Numerous seaports were distributed along the southwest coast of the peninsula. Like their eastern counterparts, these ports have been extremely important to maritime commerce throughout the historic period. Towns along the Arabian seacoast were extraordinarily cosmopolitan locales during the Vijayanagara period (and indeed long before), home to Arab, African, Jewish, Persian, and East Asian merchants, as well as diverse South Asian merchant communities.

Vijayanagara political structures Historians disagree about the nature and forms of political and administrative structures of the Vijayanagara empire, and how these varied over time and space. As summarized at the end of this chapter, three important contemporary historians of the period have each presented very different interpretations of Vijayanagara. Nilakanta Sastri saw a war state and Karashima a feudal system (for the later Vijayanagara period), while Stein viewed Vijayanagara as a segmentary state held together by ritual kingship. Recently, as noted earlier, Talbot, Wagoner, and Eaton have added still another perspective to these debates (not discussed here) by emphasizing Vijayanagara’s connections with the Bahmani state and its successors sultanates. Of


these scholars, only Karashima pays close attention to economic structures, and craft production, but each view has important implications for the study of Vijayanagara political economy. As noted above, in the early decades of Sangama expansion, Vijayanagara territories were divided among the Sangama brothers, each of whom was largely autonomous (B. Stein 1989a: 27). As the empire expanded under later Sangama rulers (i.e., Harihara II and Devaraya I and II), efforts at state centralization increased through the creation of additional administrative units. Governors, often members of the royal family, were appointed over imperial provinces: rajyas or uchavadis. Devaraya I and Devaraya II also appointed Brahman officials to administrative positions (B. Stein 1989a: 28; Nagaraju 1991: 52–58). The administrative units varied in size and inclusiveness (some large rajyas appear to have contained smaller rajyas within them); their precise number, organization, and changes over time are not well undestood. Devaraya II and his successors constructed military fortresses throughout the empire, adding a new layer of Vijayanagara authority in incorporated regions, as well as producing compelling material expressions of state power. The movement of Telugu and Kannada warriors across imperial territories began early in the Vijayanagara period, and was part of a broader pattern of demographic shifts that would have dramatic long-term consequences. However, in many areas of the empire in these early decades, local elites appear to have retained much of their authority as long as they swore fealty to the Vijayanagara ruler. This pattern would persist in some regions throughout the next century as well. Local administrators of various sorts appear to have been largely responsible for maintaining revenue records and extracting resources from the territories they governed. Karashima (1992) has argued that in Tamil regions, peasant uprisings and mass migrations in the early fifteenth century provide evidence that Vijayanagara officials had raised the exploitation of the peasantry to intolerable levels (though see Narayana Rao, Shulman and Subrahmanyam 1992: 30–31 for a critique of this argument). The best documented of these uprisings took place in AD 1426–1429 (see chapters 7 and 8). In a series of inscriptions, the protestors asserted that their actions were being carried out as a response to “the exploitative and oppressive rates of taxation that have been placed upon us by Brahmans, local rulers, the Vijayanagara military, and other imperial officers” (Karashima 1992; Morrison 2001; Sinopoli 1998) – a wide range of exploiters indeed. Whether the blame for the conflicts of the early fifteenth century can be solely laid on Vijayanagara officials, scholars have argued that a restructuring

89

90


of political organization began in the late fifteenth century in the southern regions of the empire. This shift followed the period of state decline in the late Sangama period, and parallels the beginning of the dramatic expansions of the empire under the Saluva and early Tuluva dynasties. In the southeastern part of the empire, this period is characterized by the increasing frequencies of the title nayaka in the inscriptional record. These nayakas were Telugu warriors who had been ceded rights to imperial territories in exchange for providing military forces and swearing loyalty to the Vijayanagara court. They provide evidence for a new level of powerful, non-local, political agents in this wealthy region of the empire. A considerable amount has been written about the nature and role of the nayakas and their relation with the Vijayanagara center. Karashima (1992) saw their emergence as evidence for a transition to feudal order, while Stein (1989a) viewed their appointment as an attempt at state centralization that ultimately failed. Narayana Rao, Shulman, and Subrahmanyam (1992) have presented a rather different interpretation of the arrival of nayaka lineages in the Tamil region in the early sixteenth century. They view the nayaka presence as the outgrowth of broader processes of Telugu expansion into the southern regions of the empire – “the creeping Nayaka penetration of the Tamil country” – that had begun in the early decades of the empire, if not before (Narayana Rao, Shulman, and Subrahmanyam 1992: 35). Thus, they see their movement into Tamil country as part of a broad historical process, rather than a specifically Vijayanagara phenomenon. Certainly, the nayakas’ military connections with the Vijayanagara state were an essential component of this process, but their presence may be less a Vijayanagara creation than a reflection of the Vijayanagara rulers’ acquiescence to ongoing demographic and military processes. The result was “a structure of power which derived from but was at the same time distinct from the Vijayanagara state” (Narayana Rao, Shulman, and Subrahmanyam 1992: 43). These interpretations imply very different things about the relations between nayakas and imperial rulers, and specifically how dependent nayaka rulers were on the Vijayanagara emperor. Stein and Karashima view the nayaka offices, at least initially, as being non-hereditary and instead, solely reliant on, and easily rescinded by, the Vijayanagara king. Both agreed, however, that over time, Vijayanagara became less able to control their creations. For Narayana Rao et al., nayakas were, from the beginning, members of powerful hereditary lineages who were establishing durable states within the larger Vijayanagara polity. Despite disagreements concerning the origins of the nayaka presence in Tamil regions, there is general agreement on their impact, which included the undercutting of traditional elites and


communal landholding systems, as revenue and resources were reallocated to these new “foreign” rulers. Political order and ideas of rule and kingship were also transformed in ways that accommodated very different sources of rule and authority. As the nayaka presence was increasingly felt in the southeastern regions of the empire, somewhat different administrative structures appear to have developed in its upland core. Here we see references to numerous local Kannada elites, or “poligars,” who ruled small territories and paid revenues and provided forces to the Vijayanagara kings (B. Stein 1989a: 60). More than 80 of these poligar lineages are known from family histories recorded in the early eighteenth century by British collector Thomas Munro (B. Stein 1989a: 59–60; 1989b); most traced their origins to the Vijayanagara period. Successful warriors also were granted nayaka positions and ceded territories (amaram) in upland regions (see Talbot n.d.a for a discussion of the amara-nayaka in Andhra Pradesh); these could be granted by the state and by local elites. For example, in 1563, the Alamkonda amaram (Kurnool District, Andhra Pradesh) was granted to a warrior by Rangapparajaya, a chieftain of Rayalaseema (B. Stein 1989a: 86). A record from this amaram provides one of the few revenue documents that survives from the Vijayanagara period, and is frustratingly suggestive of the kind of detailed administrative record keeping that must have existed throughout the polity. The text – the Puttasti of Alamkonda (Nilakanta Sastri and Venkataramanayya 1939, vol. III: 88–94) – dates to AD 1563, and provides information on arable land, crops, craft workshops, and their taxable revenues, spanning half a century from the time the region was first incorporated into the empire (see chapter 7, pp. 258–262). It provides striking evidence of the enormous economic expansion that occurred in the sixteenth century (as revenues from the region increased more than fourfold over less than fifty years; B. Stein 1989a: 87–88). The picture that emerges for Vijayanagara administration is extraordinarily complex, over both time and space. Indeed, the Portuguese merchant Nuniz wrote of more than 200 captains of the Vijayanagara period, who held property, raised revenues, and provided military forces (Sewell 1900: 389). While the precise titles and roles of these captains cannot be defined, it is clear that administrative officers of various sorts were widespread and numerous. Nonetheless, there is little evidence for an extensive well-established central bureaucracy; instead, much authority was delegated onto various officers and administrators who were responsible for extracting resources, providing military service, and maintaining economic records at the local level. Administrative elites included both traditional local elites, who benefited from cooperation with the empire, as well as imposed Brahman administrators,

91

92


military officers, and nayakas, among others. The results were complex “mosaics of overlapping interests and hegemonies” (B. Stein 1989a: 73). Further, as the Alamkonda example indicates, these various officers were organized at multiple hierarchical levels, creating a palimpsest of state and local authority and competing claims to power. Like many historic empires, Vijayanagara’s rulers responded strategically to local situations, rather than seeking to create an overall system of rule. Such a system would, no doubt, have been impossible to implement, and was unnecessary to meeting imperial goals of securing wealth, military supremacy, and political allegiance. It also must be remembered that political officers and office holders were not the only participants in this complex world. Temple and religious leaders, merchants, and warriors also had considerable impact on political, economic, and ideological developments. I turn to some of these impacts now, beginning with a discussion of the role of religion and temples in South India during the Vijayanagara period.

Religion in fourteenth- through seventeenth-century South India Although often portrayed as a Hindu empire and a bulwark against Islamic expansion in South India (e.g., Devi 1990; Saletore 1982), the regions ruled by Vijayanagara were also inhabited by Jains, Muslims, and practitioners of various local tribal religions. Within Hinduism itself, the Vijayanagara period produced numerous charismatic religious leaders and sectarian movements. Most important were several highly influential Shaivite and Vaishnavite devotional, or bhakti, communities, whose adherents renounced caste and priestly authority. Holy men from these communities had tremendous popular followings and served important roles as royal advisors and sources of political legitimation. Also important were the large temple institutions of South India, centers of considerable wealth and power. Vijayanagara’s kings were Hindu and donated generously to the large temples dedicated to Shiva and Vishnu. And like many Hindus, Vijayanagara’s rulers acknowledged numerous gods beyond their primary tutelary family deities. However, individual kings tended to favor particular temples, for both religious and strategic political reasons. Members of the first ruling Sangama dynasty were Shaivite. The Saluvas and Tuluvas were Vaishnavite, and particularly under the later Tuluva kings there was a shift to greater royal investment in Vaishnavite institutions, and a decline in the ecumenicalism of earlier periods (Verghese 1995, 2000). Nonetheless, throughout the empire’s history, individuals of a variety of religious traditions were employed


by the state, and religious difference was acknowledged in several juridical texts and inscriptions. While South Indian society was diverse and Vijayanagara’s rulers were not engaged in a program of religious conversion, Vijayanagara’s kings were Hindu kings – and this was important both to the practice of kingship and to royal claims to legitimacy. Thus, titles associated with the Sangama kings include: “defender of the Hindus and of varnashramadharma” (right behavior according to caste proscriptions), “protectors of the deities of the Hindu Kings,” and “establishers of the Vedic path,” among others (Kulke 1985; B. Stein 1980: 369). All of these, along with several of the dynasty’s origin stories (Subrahmanyam 1998c, Sinopoli 2000), attest to the status of the Sangamas as rulers whose kingship was sanctified by gods and legitimized by holy men. Other royal titles attributed to the first Sangama ruler had rather different associations. In an inscription dating to AD 1352, Bukka I was termed “the prosperous great tributary, punisher of enemy kings, Sultan among Hindu Kings (Hindu-raya-suratrana), vanquisher of kings who break their word, lord of the eastern and western oceans, the auspicious hero” (Wagoner 1996: 861). Wagoner (1996) has convincingly argued that the title Sultan among Hindu Kings provides evidence that Vijayanagara’s kings participated in a pan-South Indian political culture that also incorporated Islamic political traditions (also Sinopoli 2000). The subjects of the Vijayanagara state, as noted above, worshipped many deities, including local goddesses and powerful natural sites, regional deities, and gods affiliated with particular occupational communities or kin groups, Vaishnavite and Shaivite deities, along with Jaina saints, and Allah and Muslim saints. The countless small shrines and large temples found throughout the Vijayanagara metropolitan region (Sinopoli 1993a; chapter 5, this text), provide evidence for the diversity of South Indian religious practices and institutions, which varied widely in scale and in sponsorship. As I elaborate in chapters 6 and 7, a broad range of craft producers were involved in producing sacred images, structures, and other materials used in worship. These producers ranged from rural carpenters and stone carvers to elite poets and bards who were based in the large, elaborate temple complexes. The popularity and status of individual gods also changed over time. Throughout the Vijayanagara period, certain local deities, particularly those associated with the Vijayanagara military, experienced an elevation in their ritual status, and were transformed (“Sanskritized”) from regional deities to manifestations of Vishnu or Shiva, whose temples were the recipients of royal donations. Like political structures discussed above, religion in South India was dynamic and complex, and individuals simultaneously participated in

93

94


multiple religious communities. Further, although the king and court was actively involved in religion, through the sponsorship of gods and temples, there was no state religion in the sense of royal control of religious institutions or practices. Instead, royalty and other elites vied for the support of temples and religious leaders, adding another layer of complexity and competition to the multifaceted political dynamics of the period. During the fourteenth through seventeenth centuries, many major temple centers of South India grew to urban proportions, as large settlements grew up around them, and they became centers of commerce and production. The focus of worship and pilgrimage, these large temple complexes were also the focus of elite political competition among kings and nobles, and were centers of enormous wealth. Many of the donations to temples were reinvested in agriculture. Temple centers were also involved in the expansion of craft production and commerce, particularly in textiles and metalworking. Below, I summarize several features of the large temple centers, although I reserve the discussion of their relations to craft production and producers until chapters 6 and 7.

The South Indian temple Much of what scholars know about South Indian history derives from the analysis of inscriptions carved on the walls of temples, and particularly the large temple complexes that were the major recipients of elite donations. In part as a result, the role of the temple has figured importantly in historical interpretations. The development of temples into major economic and political forces, as well as sacred institutions, began in South India several centuries before the rise of Vijayanagara (e.g., Heitzman 1987, 1997), but the process greatly accelerated with the expanded wealth, scale, and political competition of the Vijayanagara period. The best-documented and largest temple centers of the period are located in Tamil Nadu and southern Andhra Pradesh, areas far from the imperial capital (although as discussed in chapter 5, the city of Vijayanagara also contained major temples and pilgrimage sites). And it is in these areas particularly that temples became important loci of economic and urban expansion. The large pilgrimage centers of South India, such as Tirupati, Tiruvannamalai, Kanchipuram, and Srirangam, among others, underwent dramatic growth during the Vijayanagara period. This growth, as measured by the size and number of temple donations, peaked in the early sixteenth-century Tuluva period. Prominent donors included royalty, local elites, nayakas, military leaders, and merchants, among others. Temple expansion is also evident in architectural changes


4.7 Jalakantesvhara Temple, Vellore

at the temple centers (Michell 1995), which grew in size and scale through the addition of imposing towered gateways and massive enclosure walls (Figure 4.7). Arjun Appadurai and Carol Breckenridge have written extensively about the role of South Indian temples in mediating relations between kings, religious sectarian leaders, local elites, and deities (Appadurai 1977, 1981; Appadurai and Breckenridge 1976; Breckenridge 1986). Temple donations

95

96


became a focus for political competition among diverse elites, as well as expressions of fealty and devotion. Donors received both prestige and economic benefits, and detailed records of their donations and their distribution were inscribed on the walls of structures and on copper plates. Temple administrators and sectarian religious leaders became important political actors in their own right, whose support was crucial to the various elites who sought them out. Major temples were also centers of devotion and social life for thousands of pilgrims – elites and commoners – who participated in religious worship and economic transactions at increasing numbers of calendrical and occasional festivals (Mack 2000, 2002). They were foci for population movement, communication, and propaganda. Temples were also mediators of a range of legal disputes, and the claims and resolutions of these disputes were also recorded in inscriptions on their walls. South Indian temples figured importantly in processes of economic expansion during the Vijayanagara period. Donations of land and wealth contributed to agricultural expansion. Temple donations were deployed in the construction of irrigation works and the transformation of waste or forested regions to agricultural zones (e.g., Soma Reddy 1984). Temples thus were also involved in the period’s large-scale demographic shifts, as farming communities and artisans expanded into these new territories. Most significant for present purposes, temples played a major role in the expansion of craft production and urbanism. As noted above, many temple centers became centers of population growth as small cities grew around the temple nucleus. The residents of these cities include the hundreds (in some cases, thousands) of people who served the temple itself: priests, musicians, poets, accountants, dancers (Parasher and Naik 1987), sweepers, carpenters, masons, potters, smiths, and weavers, among others (Nagaswamy 1965). Many temple centers and their elite patrons sought to attract additional craft producers to their environs. Artisans, particularly weavers and smiths, were lured to temple towns through promises of tax abatements, and towns such as Kanchipuram, Tiruvannamalai, and Madurai became major centers of commercial production and trade. At the same time, as I discuss in chapters 6 and 7, certain artisans and artisan communities attained considerable wealth, and were granted important roles in temple administration. Hindu temples were major participants in the social, economic, and political developments of fourteenth- through seventeenth-century South India. As centers of worship, faith, and wealth, the influence of major temples extended over vast areas, attracting donors and pilgrims from throughout (and beyond) imperial territories. As a focus of political competition, temples were crucial to the legitimation of kings and local rulers. With regards to


craft production, large temples were major consumers of a range of goods. Poets and musicians composed tributes and sacred songs in honor of the gods, and some gained considerable wealth and prominence as a result. Weavers produced the elaborate garments worn by sacred images and temple functionaries; jewelers produced ornaments of gold, silver, and precious stones to adorn images and structures; carpenters, masons, and sculptors shaped images and permanent and temporary structures to house them. The large temples fed thousands of pilgrims, and required numerous brass and earthenware vessels to stock their kitchens and feeding houses. The economic vitality of the large temple centers is also evident in the markets and craft workshops that arose around them. While artisans were rarely temple donors and do not appear often in the inscriptional record, the inscriptions that do occur provide a useful source of information concerning the relations between temples and craft producers (chapters 5 and 7).

Society and economy In this section, I address several issues concerning South Indian social and economic structures in the fourteenth through seventeenth centuries AD. I first consider South Indian social structures. Rather than recount the extensive literature on the nature and significance of caste, I restrict my discussion to a consideration of the constitution of occupational communities. Craft-producing communities were typically low in both social and economic status, and were members of local hereditary groups: lineages or subcastes. Despite this, there was tremendous social mobility throughout the Vijayanagara period, as individuals and communities benefited from expanding economic and military opportunities and the opening up of new areas of settlement. After a brief discussion of the constitution of individual communities, I also address inter-caste relations at the local level (the ayagar or village servants) and the existence of broader social groupings and affiliations in historic South India. The prominence of such groups as the valangai and idangai (right- and left-hand divisions) dramatically undercuts the views of Mill, Marx, and Weber, as well as many proponents of Indian feudalism and the Indian historiographical approach (see chapter 3) concerning isolated village India. As I discuss here and in chapter 7, these groups provide evidence of large-scale social identities and social action involving multiple social units over broad geographic areas. I then turn to three issues on concerning historic South Indian economic structures that are specifically relevant to the study of craft production:

97

98


trade associations, monetization, and labor recruitment. Trade associations or “guilds” have a long history in South India. The largest of these associations, the Ayyavole 500, extended over much of Karnataka, Andhra Pradesh, and Tamil Nadu during the eleventh through fourteenth centuries AD. Although these groups were linked to state expansion, they provided another autonomous dimension to processes of economic and urban growth in South India. A second important trend during the Vijayanagara period was the monetization of the economy. Over time, cash donations became increasingly important in temples and as a medium for taxation. Local craft producers and agriculturalists were increasingly drawn into a monetized economy in ways that dramatically affected their lives. Third, I consider what is known about the recruitment of large-scale work groups during the Vijayanagara period. Despite the material evidence for massive constructions and public works during the Vijayanagara period, we know relatively little about how labor was recruited and organized for major construction projects. Several sources suggest that enormous numbers of workers could be mobilized for projects such as reservoir construction; these groups provide a marked contrast to the more widespread pattern of household-level craft production that I discuss in chapter 6.

Occupational groups: the problem of caste Historic South India was characterized by tremendous differentiation along multiple dimensions – including geography, political history and structures, language, and religion. Accompanying this was enormous horizontal and vertical social differentiation. In this section, I focus on Vijayanagara society and in particular on occupational differentiation and other issues relevant to understanding who craft producers were and where they fit in their broader society. The subject of the South Asian caste “system” has, for much of the last century, dominated scholarship on South Asian social organization, which has focused on understanding the hierarchical division of the social universe into endogamous groups who shared religious affiliation, origin stories, and social practices and proscriptions (e.g., dietary proscriptions), along with their primary occupation(s). And caste as a social category is certainly an important component to any understanding of Vijayanagara craft production. However, documenting what “caste” means in terms of behaviorally meaningful social and occupational groups has proven far more difficult than the study of its idealized structure as described in sacred texts (e.g., see for example, Appadurai 1986; Bayly 1999; Dirks 1988, 2001; Inden 1990). Terms such as gotra (lineage), subcaste (jati), caste-clusters, and varna (the four


divisions of Brahman, Kshatriya, Vaishya, and Shudra; although, the middle two categories are largely absent in South India), have been used to capture increasingly inclusive levels of relationships and social identity, from the local to the universal. Talbot’s analysis of inscriptions from Kakatiya through Vijayanagara-period Andhra Pradesh suggests an even more complicated picture in which occupational identity, an important component of caste, also played a major role in the construction of social identities in ways that crosscut heredity. Thus, she notes, the title setti was used to identify diverse merchant and artisan communities across the South, while boya was used as a generic title that indicated herders (Talbot 2001: 58–59). Talbot argues that the Kakatiya social universe was indeed divided into categories or “discrete functional orders” as a caste model would suggest. But, she suggests, these orders were based primarily on “the centrality of occupation,” rather than more abstract constructs defined by inheritance and position in a sacred hierarchy (Talbot 2001: 61). The behavioral flexibility, regional variability, and high degree of social mobility that characterized historic South Indian society makes it problematic to use caste terminology to characterize artisan groups. While it is possible in some sense to speak of a “potters’ caste” (or jati) or “smith caste” at a pan-South Asian scale, such categories had little behavioral meaning for many artisans, whose economic and social interactions often (though not always) played out at more localized scales. Regional endogamous “subcastes” or lineages were more limited communities of producers, operative at the level of individual settlement or cluster of associated settlements. These groups acted as corporate units in coordinating tax payments, regulating production and behavior, settling disputes, and organizing religious festivals, among other activities. However, as discussed here and in chapter 7, in some contexts members of multiple occupational groups acted corporately during the Vijayanagara period (e.g., the right- and left-hand castes). Further complicating the issue, occupational titles that appear in inscriptions may, as noted above, record social identity, but need not in all cases imply kin-based affiliation, and thus may or may not be caste titles. While I will not entirely discard the language of caste, rather than classifying relevant groups of craft producers into problematic categories of castes, subcastes, or the like, I follow other contemporary South Asianists in using the vaguer and more elastic term “community” to refer to behaviorally meaningful social collectives of craft producers, without constraining the constitution of these groups too narrowly or in ways unsupported by available evidence. Within occupational communities, the specific tasks that individuals performed varied with sex, age, and skill, as well as with wealth and individual status. Among potters, as noted in chapter 2, only males traditionally used the

99

100


potter’s wheel, though women performed numerous other tasks associated with pottery production. And not all members of a hereditary community necessarily engaged in their “traditional” occupation. During the Chola period, many members of the prominent and widespread Kaikkolar weaving community were employed as soldiers, an occupation that largely disappeared among Kaikkolar by the fourteenth century, when many shepherds took up military positions (Mines 1984). During the Vijayanagara period several Kaikkolar individuals emerged as master weavers, individuals who owned or supervised many looms (see chapter 6, pp. 187–190). It is unlikely that such individuals engaged in the physical labor of weaving. Some occupations were not exclusively restricted to members of specific hereditary groups; for example, in cotton-growing regions the majority of non-elite women engaged in spinning, no matter what their traditional hereditary occupation. The Vijayanagara period, like other periods of Indian history, was a time of population mobility and at least some degree of occupational flexibility. A person’s hereditary occupation was extremely important, but it was by no means the sole determinant of an individual’s occupation or social position. As noted, most producers of material crafts were of relatively low, nonelite, social and economic status (I exclude poets and bards from this discussion, as these individuals came from social groups ranging from Brahmans and royalty to shepherds; see chapter 6). The position of craft producers in regional social hierarchies varied along a number of dimensions. These included the social and political valuation attributed to craft products, variables that could change over time. Potters, producers of goods with low social value, also had low social and economic status throughout the Vijayanagara period. Other artisans who produced more esteemed goods, such as some kinds of textiles and metal objects, were able to improve their status over the course of the Vijayanagara period. Dietary practices (e.g., vegetarian or non-vegetarian), technology and materials employed in the craft, marriage patterns, religious affiliation and practices (e.g., practice of animal sacrifice), and a range of other behaviors and traditions also constituted and contributed to the status of particular artisan communities.

Social relations at community and regional levels Ayagar or village servants The high degree of economic specialization that characterized historic South Indian society necessarily implies a similarly high degree of economic interdependence. Elites and non-elites alike were mutually implicated, in the


need for basic subsistence and material products, and for ritual and other services. While, during the Vijayanagara period, a cash economy provided one means for the exchange of goods and services, long-standing obligatory relations and inter-household exchanges provided other means for the fulfillment of these needs. The internal organization of villages, and perhaps districts within larger urban settlements, also supported a category of people known as ayagar, or village servants. The ayagar were families or performers of occupations deemed essential to the inhabitants of a settlement, who received payment for serving the community. References to ayagars appear in several Vijayanagara-period inscriptions. They also appear in a number of documents in the Mackenzie Collection, a set of local histories recorded by Colin Mackenzie and his associates around 1800. Mackenzie was a military surveyor in southern India, with antiquarian interests and tremendous energy. Working with local pundits and scholars throughout the south, he recorded numerous inscriptions and local histories (kafiyats) from regional polities throughout South India (Mahalingam 1972). Several of these texts were excerpted by Nilakanta Sastri and Venkataramanayya (1939) and provide information on ayagars. For example, a text called the Atthavanavyaharatantram (Mackenzie Mss. 15-6-8, Sec. 10, pp. 3–4) notes: There are twelve functionaries (or officers) for every village: 1. Accountant 7. Shoe-maker 2. Headman 8. Goldsmith 3. Carpenter 9. Watchman 4. Washerman 10. Waterman 5. Purohit [priest] 11. Black-smith 6. Barber 12. Potter

(Nilakanta Sastri and Venkataramanayya 1939: 310–313)

These ayagars are described as receiving shares of agricultural produce as well as payments from taxes paid to the village headmen. The text continues: The carpenter and the blacksmith attend to all artisan’s work in the village, especially to the manufacture of ploughs and other agricultural implements, without demanding wages for their services . . . The cobbler furnishes the ryots [peasants] with ropes, buckets, sandals, etc . . . Every village does not maintain a separate potter. One or two potters supply the pots necessary for use of all the ryots of a taraf (district). However, the potter has an ayam [income] in every village. He is also in the habit of selling pots in the bazaar, for which he has to pay a tax called cakrakanike to the government . . . These twelve ayagars are attached to the village; and their offices are hereditary.

(Nilakanta Sastri and Venkataramanayya 1939: 310–313)

101

102


A second Mackenzie document (Nilakanta Sastri and Venkataramanayya 1939: 313–314) lists a somewhat different set of ayagars: 1. Purohit (village priest), 2. Lekhaka (accountant), Devalakdadvayam (arcakas of the Saiva and Vaishnavite shrines), 4. Ayakharah (blacksmith), 5. Vardhaki (carpenter), 6. Ksuri (barber), 7. Rajakah (washerman), 8. Vadajah (protector of the village boundary), 9. Padvaijneyah (knower of the foot-prints), 10. Antayjatidvayam (described as “the two of the last caste”: e.g., shoemaker, washerman, or other polluting, low-status occupation). Potters and goldsmiths are absent from this group. While the composition of occupations classified as ayagar or village servants varied somewhat, these individuals were responsible for providing necessary goods and services to village residents. It is likely that during Vijayanagara times, particularly in smaller rural settlements, some portion of artisans were engaged in these types of relations, although they may have also, as the potters discussed above, participated in market economies and a variety of other exchange relations.

Right- and left-hand divisions As noted above, while many meaningful craft-producing community groups were restricted to relatively small and geographically limited regions, there is evidence for the existence of larger social units that transcended caste and occupational boundaries. One such grouping appeared in the Tamil region at around AD 1000 (Appadurai 1974) and is known from southern Karnataka at approximately the same time. The valangai and idangai (Tamil; balagey and edagey, Kannada), or right- and left-hand divisions, were groups that included members of a large number of relatively low-status local subcastes (conventionally, 98 right-hand and 98 left-hand castes are described as comprising these divisions). In a general sense, the right-hand division included groups traditionally involved in agriculture or the production of agricultural or rural commodities, though it also had associations with military service. Communities belonging to the left-hand division were primarily non-agricultural, and included merchant and artisan groups (B. Stein 1980, Beck 1970). Although this suggests a rural–urban distinction, the criteria for assigning right- and left-hand status are not clear for some groups. For example, the Kaikkolar weavers belonged to the left-hand division, while the Saliyar weaving community belonged to the right. Burton Stein (1980: 176) has suggested that these horizontal social divisions became important in processes of assimilation of new communities into the expansionist political economies of tenth- through


fourteenth-century South India. During this time of economic and urban growth, these structures provided a framework to incorporate into broader political and economic spheres groups that had been traditionally outside of caste and state hierarchies. Several Chola-period inscriptions refer to formal granting, in the presence of local elites and Brahmans, of idangai or left-hand status to specific named communities (B. Stein 1980: 182–184), providing evidence that that group was actively expanding its membership during this period. Affiliation with valangai or idangai divisions was most important to identity and social action during the pre- and early Vijayanagara periods, particularly in Tamil regions. Material insignia of affiliation appeared; the right-hand division was associated with the elephant, eagle, and half humanhalf animal form, while the left-hand division was associated with the fox, tiger, horse, lion, and yali mythological animal (B. Stein 1980: 185). During the Chola period, the valangai and idangai appear as donors to temples, as well as litigants in tax disputes and disputes concerning temple or privileges. Texts and inscriptions provide evidence for competition and conflicts between right- and left-hand divisions. However, in other contexts the two groups effectively united, as in a tax protest of the early 1420s in which the 98 right-hand and 98 left-hand castes joined together and successfully protested oppressive Vijayanagara tax rates (Karashima 1992; also chapter 7, this volume). Although the terms valangai and idangai remained in use for more than eight centuries, they declined in frequency in the later Vijayanagara period. This appears to reflect their dwindling significance as meaningful and effective corporate groups. It may be that, as Stein suggested, the major importance of these groups as effective units occurred during the most dramatic periods of urban expansion and economic transformations, beginning in the Chola period and extending through the first century of the Vijayanagara period.

Trade associations Trade associations, often referred to as guilds, appeared in the South Indian historical record from the eighth century AD (Champakalakshmi 1996: 313; Abraham 1988). Several such associations are documented through the Vijayanagara period, though their significance and structure seems to have changed over time. These groups varied in scale and focus – from cooperative associations of merchants based at a single urban settlement (nagaram), to associations involved with a single commodity such as textiles (e.g., the

103

104


“Mahandau” association of Kanchipuram), to larger super-regional merchant associations. Most important among the latter was a group known as the Ayyavole 500. Inscriptions referring to this group first appear in the Krishna-Tungabhadra region of Karnataka in the eighth century at the site of Aihole. From there, the group expanded to Tamil Nadu and Andhra Pradesh (Abraham 1988). Other somewhat more localized associations include the “Manigaram,” who first appear in Kerala in the ninth century and were particularly active in maritime trade. The Manigaram had largely disappeared from the inscriptional record by the fourteenth century. Other named groups include the Valanjiyar, Nanadesi, Cittarameli, and Anjuvannam; these were more localized or had more restricted membership than the Ayyavole 500 or Manigaram (Champakalakshmi 1996: 311). The Anjuvannam, for example, was an association of foreign merchants (Jewish and Arab) based primarily in Kerala (Champakalakshmi 1996: 313). The Ayyavole 500 was most the most extensive and important of the large trade associations prominent prior to the Vijayanagara period. It continued to play a role in regulating inland and overseas commerce through the seventeenth century. Champakalakshmi (1996: 316) described the Ayyavole 500 as “a group of people of disparate origins associating together for a common purpose (trade).” Individuals of multiple castes, religions, and geographic origins participated in this loosely connected association of itinerant merchants. The expansion of this association in Tamil Nadu paralleled the expansion of Chola hegemony; although not controlled by the state, commerce and merchant guilds benefited from state expansion and formed close associations with dynastic elites (Abraham 1988: 75). Merchants of the Ayyavole 500 established close links with agricultural communities and craft producers (Abraham 1988: 81). They also contributed to, and benefited from, broader processes of urban expansion; an inscription dated AD 1267 from the gold-rich regions of southern Karnataka referred to a member of the Ayyavole guild as a “maker of towns” (Abraham 1988: 64). The Ayyavole 500 was engaged in facilitating and protecting long-distance trade – both inland and maritime. They employed mercenaries to protect merchant caravans and established fortified towns and warehouses to secure their goods. A wide range of commodities appears in their inscriptions, including foodstuffs, horses, elephants, and textiles, as well as goods destined for royal courts, including silk, sandalwood, perfume, pearls, sapphires, rubies, lapis lazuli, and gold (Champakalakshmi 1996: 523–525). The role of the trade associations seems to have changed significantly during the Vijayanagara period. Some associations, such as the Manigaram, disappeared completely, while the focus of the Ayyavole 500 became more


geographically restricted. Most Vijayanagara-period references to Ayyavole 500 merchants occur in the cotton-growing districts of Andhra Pradesh; they disappear completely in Tamil Nadu. While these changes have yet to be studied in detail, Chamapakalakshmi (1996) has suggested several contributing factors, including the increased monetization of the economy. She also suggests that the increasing prominence of individual traders, master craftsmen, and local elites played a role in usurping the authority of previously dominant itinerant trading communities.

Monetization To the left of the king’s court is the mint. They have three types of gold alloy. One is called varaha . . . The second is called partab, half of the former. The third is called fanam, a tenth of the [varaha]. The coin most current is the fanam. They make a sixth of a fanam from pure silver and call it a tar. It too is much in currency. There is a third of a tar made of brass; it is called a chital. (Abdul Razzaq in Thackston

1989: 309)

So the Timurid ambassador Abdul Razzaq described the mint and coinage he observed during his visit to the Vijayanagara capital in the mid-1400s. He continued: “It is the custom of the realm for all of the provinces to bring gold to the mint at an established time . . . The soldiers receive their salaries every four months” (in Thackston 1989: 309). He thus presents a picture of a highly monetized economy by the fifteenth century. The high frequency of cash donations to large temples such as those at Tirupati/Tirumalai (Mack 2000) from the Sangama period on supports the idea that currency was both widespread and widely used. As evident from Razzaq’s description, Vijayanagara-period coinage included gold, silver, and copper standards. Coinage was minted by the state, as well as by regional centers. Merchant guilds also minted coins, and foreign coins circulated throughout the peninsula, and were valued by weight and material. Of South Indian coins, the gold varaha (also hun or gadyana; called pagoda by the Portuguese) served as the standard high value coinage, used in commerce, taxes, and temple donations. The varaha ranged from 50.65 to 53.00 grains (with a standard of 52 derived from the seed of the kalanju or Molluca bean; Appadorai 1990 [1936]: 709, Subrahmanyam 1998b: 194). The panam coin, an alloy of gold and copper, was produced in many areas of southern India and varied considerably in value. These coins were most important in local-level commercial transactions (Subrahmanyam 1998b: 194). Silver coinage was relatively rare in southern India. In contrast,

105

106


low-value copper coins (the kasu and nevel) were very widespread, and copper was a significant import into southern ports in the sixteenth century, and probably earlier (Subrahmanyam 1998b: 195). The use of coinage in trade, taxation, and temple donations evident during Vijayanagara times had a long history in South India, and was certainly not an innovation of the Vijayanagara period. However, there is evidence that the use of money increased considerably as a medium of commerce, payment, and taxation (Subrahmanyam 1990a; B. Stein 1989a). Burton Stein (1989a: 73–74) attributed this increase primarily to expanding international trade, in which Indian merchants typically received currency in exchange for a range of commodities including textiles, metals, and spices (the exception to this pattern was the trade in horses, which involved export of bullion from India; B. Stein 1989a). Other factors contributing to this development include the increase in internal commerce and commodity production at urban centers, and overall patterns of population growth and economic (including agrarian) expansion (Subrahmanyam 1990a, 1998a). Subrahmanyam (1990a: 357) has suggested that while the state was not the direct cause of expanding monetization, it did benefit from it, in particularly from the improved record keeping that must have accompanied commercial activities and the ability to better track revenues and resources. Further, the collection of taxes in money rather than in kind, appears to have increased dramatically throughout the empire’s history; in fact, B. Stein (1989a: 64) suggests all taxes came to be paid in currency by the sixteenth century. This was a marked contrast to earlier periods. References to cash taxes were rare in Chola sources, but by the sixteenth century, ten categories of money taxes appear to have been collected in these regions. These included taxes on agriculture and herding, on forest products, on imports, and professional taxes paid by barbers, leather workers, hunters, and other occupational groups, and marriage taxes (see chapter 7). Along with commerce and taxation, currency also circulated via temple donations. However, here patterns of donation in currency, land, or goods appear to have varied widely. At the Tirupati and Tirumalai temples, for example, money was the most common donation throughout the Vijayanagara period, though it declined slightly in importance over time. In contrast, at Tiruvannamalai, cash donations were rare and donations of land constituted more than 70 percent of all Vijayanagara-period gifts (Mack 2000). As a whole, however, the general pattern across South India was of expanding monetization. While overall trends suggest relations between monetization and expansion in commerce and agrarian and craft production, the causal links


between these processes are difficult to determine. International and internal trade, political needs (such as the support of a standing military), temple investments, and general patterns of population growth and movement, all no doubt played a role. Whatever the ultimate causes, the period was one of market expansion, and the lives of those who produced products for markets – including many craft producers – no doubt changed significantly as well, as they were increasingly drawn into monetary economies.

Labor recruitment Kinship, region, and hereditary occupation clearly played major roles in structuring labor organization in the Vijayanagara times, and as discussed in chapter 6, production at the household level was dominant for many kinds of labor. However, there is evidence for the existence of much larger-scale production units, although we know disappointingly little about how such groups were recruited or compensated. Texts sometimes refer to work groups involved in the construction of irrigation works or engaged in military activities. Some of these sources provide evidence for enormous numbers of laborers engaged in production activities for extended periods. For example, Domingo Paes described the construction of an earth and masonry reservoir embankment on the outskirts of the Vijayanagara imperial capital: In order to make this tank the said king broke down a hill which enclosed the ground occupied by the said tank. In the tank I saw so many people at work that there must have been fifteen or twenty thousand men, looking like ants, so that you could not see the ground on which they walked, so many there were; this tank the king portioned out amongst his captains, each of whom had the duty of seeing that the people placed under him did their work, and that the tank was finished and brought to completion. (quoted in Sewell 1900: 244–245)

An inscription from Cuddapah district (Andhra Pradesh), dated to AD 1369, is unique in the wealth of detail it provides on the construction of a similar reservoir embankment (see Ramaswamy 1993; Morrison 1995). The inscription describes the construction of a massive reservoir (the Porumamilla reservoir) sponsored by Bhaskara Bhavadure, a son of the Vijayanagara ruler Bukka and governor of the administrative district of Udayagiri. The reservoir is reported as being approximately six kilometers long, seven meters high, and nine meters wide. The inscription notes that successful construction of the reservoir requires a “righteous king and a Brahamn well versed in hydraulics . . . [and a] gang of men skilled in tank construction” (Ramaswamy 1993: 30; Epigraphia Indica vol. XIV, no. 4), and records that

107

108


the embankment was constructed over two years by a thousand laborers. The use of a hundred carts from the “masonry work” is also noted. While we thus have convincing evidence for the employment of large numbers of workers in construction tasks, we know little about how these workers were recruited. Paes’ mention of work groups under the control of individual captains is suggestive, but it is unclear how accurate his perceptions were. Interpretations of Vijayanagara-period records as regards issues of slavery, serfdom, wage labor, access to and ownership of resources, etc., are typically closely bound to the theoretical predispositions of the authors presenting them, and the sources themselves often provide relatively little support to grounds to choose between differing perspectives. This becomes apparent, for example, in discussions of the subject of forced resettlement, a practice known from many imperial societies. Numerous inscriptions refer, for example, to communities of weavers or smiths being “settled” on temple lands or in new communities. Some scholars, such as Karashima (1992 and below) explicitly interpret this as evidence for coercive resettlement by state or regional leaders. For others, such as Ramaswamy (1985c) and Burton Stein (1989a), the mobility of artisans is interpreted as evidence of their ability to affect their relations with diverse institutions in order to improve their economic and social conditions. The interpretive differences rest on how scholars read the inscriptions and interpret or translate individual terms, and this often rests in the theoretical presuppositions that influence those readings. I am thus not sure that more “careful” readings could help resolve this interpretive challenge. Instead, it is important to be aware that written texts are by no means neutral sources, and that, particularly when one relies on secondary sources and translations – as I do in this study – additional interpretive layers are added on to any reading. Stephen (1997: 50) has suggested that data from the Coromandel region of Tamil Nadu indicate that peasant labor for agricultural tasks was recruited through multiple means, including bonded labor, free labor, owner cultivators, and landless cultivators. It is likely that large-scale production units, such as those discussed above, also involved a variety of labor categories, though at present we can say little more about them.

Three models of the Vijayanagara state: Nilakanti Sastri, Stein, Karashima Three historians have figured prominently in recent discussions of the Vijayanagara state. These scholars belong to three distinct traditions of


contemporary Indian historiography (see chapter 3). K.A. Nilakanta Sastri can be grouped with the Indian historiographical tradition; Burton Stein was (primarily) a proponent of the segmentary state approach; and Noboru Karashima uses a feudal framework for understanding Vijayanagara organization. Together the writings of these scholars have largely defined the parameters for historical, and to a lesser extent archaeological, research on Vijayanagara from the 1950s to the present. Of these scholars, only Karashima has written specifically about craft producers. However, both Nilakanta Sastri and Stein noted broad economic changes of the Vijayanagara period, and make reference to the social and political contexts of production. In this section, I briefly outline their interpretations of the Vijayanagara polity, and particularly how they affect interpretations of Vijayanagara craft production.

K.A. Nilakanta Sastri: the Indian historiographical tradition As outlined in chapter 3, proponents of the Indian historiographical tradition argue that, far from being the stagnant despotic polities that many European scholars believed, traditional Indian states (of the pre- and postmedieval periods) were characterized by strong kings and well-organized, efficient, administrative structures. K.A. Nilakanta Sastri, a leading figure in this tradition, argued that in southern India and under the best conditions, such state structures also “combined vigorous central control with a very large measure of local autonomy” (1966: 174) – merging the strong king with the independent village. Nilakanta Sastri did not author a major work on Vijayanagara. Instead, his views of the empire are contained within his classic A History of South India, first published in 1955, and revised subsequently (here I refer to the third edition of this work published in 1966). This comprehensive book presents a history of South India “from prehistoric times through the Vijayanagara period.” A major goal of the author was to bring the South into broader discussions of Indian history and civilization, which had until that time had a strongly northern focus. His work was not theoretically radical; however, it did provide solid scholarly support for subsequent discussions of South Indian history. Nilakanta Sastri viewed the Chola empire as the apogee of South Indian statehood. The Cholas ruled “an extensive and well-knit empire efficiently organized and administered, rich in resources, and possessed of a powerful standing army” (1966: 180). Importantly, the Cholas also supported local communal institutions. The result was “a wonderful social harmony” (1966: 206), accommodating the great and small alike. During the Chola period the

109

110


arts flourished in temples, palaces, and village contexts; elaborate temples were built; and the economy expanded. Nilakanta Sastri greatly admired the accomplishments of Vijayanagara. However, he saw Vijayanagara as very unlike the Cholas. Vijayanagara was a state that ultimately failed to live up to the Chola ideal due to the particular circumstances of its emergence and its mission. That mission was no less than the preservation of Hindu culture in the face of external threats, specifically the threat of Islam. Because of this, Vijayanagara was in essence a “war-state” with its structures and institutions dominated by military needs (1966: 308). As a result, some ideals of South Indian statehood were necessarily sacrificed. Indeed, Nilakanta Sastri’s history of Vijayanagara is largely a recounting of battles, of victories and defeats against internal and external foes. Despite the fact that many of the wars he recounts were between Vijayanagara and various Hindu polities of the South, for Nilakanta Sastri it was the great conflict between the forces of Islam and the purity of Hindu civilization that gave Vijayanagara its reason for existence, and its grandeur. Nilakanta Sastri argued that Vijayanagara was “best looked upon as a confederacy of many chieftains co-operating under the leadership of the biggest among them” (1996: 9). As long as local rulers obliged by providing military forces and taxes, Vijayanagara was little involved in local administration, content to let traditional organizations and administrative practices continue. However, over time changes in local institutions did occur, particularly in Tamil regions, in response to the extensive military demands of the state. And ultimately, Vijayanagara’s martial priorities led to the weakening of South India’s (i.e., Tamil Nadu’s) “most valued civil institutions,” of local autonomy and communal village governance (Nilakanta Sastri 1966: 11, also 308). In addition, because of its epic military mission, Vijayanagara’s rulers had to further compromise South India’s purity by recruiting foreign, Muslim, soldiers into its military, and by forming occasional alliances with Muslim rulers. Despite these shortcomings, which ultimately laid the grounds for the empire’s collapse, Nilakanta Sastri viewed Vijayanagara as a high point in South Indian history, particularly during the reign of Krishnadevaraya. When the king was strong, he argued, the empire as a whole was characterized by a wellorganized bureaucracy that carefully monitored revenues and trade (1966: 306–309) – it could approach the ideal Hindu state. However, the state’s main priorities were limited to the provisioning of the royal household and the maintenance of the military. The political economy of the Vijayanagara empire was dominated by its military focus, with endowments to temples constituting the other major focus for the expenditure of imperial resources.


The system worked well under strong rulers, but when kings were weak, the many warrior chiefs on whom the emperor was forced to rely could pose significant challenges to imperial unity. Nilakanta Sastri’s discussion of economic matters is far less detailed than his attention to political history, and he provides only a very general view of South Indian economy summarized from sources that span the tenth through nineteenth centuries. He emphasizes the universality of caste “with all its social and economic implications” in “upholding the social order” (1966: 319), though notes, without explanation, the “curious instances” when some artisans were formally acknowledged or rewarded by kings or temples. Most craft production and agricultural activities, however, occurred in village contexts, particularly during the Chola period, and “the rule was production for the local market” (Nilakanta Sastri 1966: 319). Presumably, the exception to this lay in goods needed to serve the military, over which the state is believed to have exercised effective centralized control. In addition, while not specifically addressing craft production, Nilakanta Sastri did argue that the autonomy of the village was undercut because of the Vijayanagara state’s military mission, resulting in the loss of traditional freedoms for non-elites, and a decline in the quality of the arts.

Burton Stein: the segmentary state Burton Stein’s 1980 publication Peasant State and Society in South India was a landmark in South Indian historiography. As noted in chapter 3, while Nilakanta Sastri had elevated South Indian historiography to national attention, Stein’s controversial work was the first to place South India at the center of theoretical debates on the nature of the Indian state. It has had a tremendous impact on scholarship in the region, generating numerous thoughtful works by its supporters and its critics. The major focus of the 1980 book was on the Chola empire, though like Nilakanta Sastri, Stein devoted a lengthy chapter to Vijayanagara. His thoughts on Vijayanagara were expanded upon in subsequent works (1985, 1989a, 1995), with some significant changes as he struggled to accommodate the unique characteristics of Vijayanagara within his segmentary state framework. Here I focus primarily on his later writings, particularly the 1989 book Vijayanagara and his 1995 review article on the segmentary state. For Stein the value of the segmentary state concept was as a structural framework that could allow scholars to evaluate the relationships among parts or segments of a political structure. As discussed in chapter 3, Stein

111

112


(1995: 137–139) argued that a segmentary state is composed of numerous coexisting centers or political domains, among which power and sovereignty are distributed. Each of these domains has autonomous administrative and coercive capabilities. These units are linked into a single polity through their shared acknowledgement of the ritual and royal authority of an anointed king who is the state’s symbolic center, based at its main ritual center. The result is a pyramidal social order, with a high degree of structural redundancy among ranked segments. The model does not specify the precise content or form of administrative, military, economic, or sacred organization, which can in any case be expected to vary from state to state. Stein argued that Vijayanagara qualified as a segmentary state because power was distributed among diverse segments. However, although local groups and rights remained important, the significant segments for Vijayanagara were different than the local assemblies and territorial units of the earlier lowland Cholas. In Vijayanagara, they also included domains ruled by military elites, which had been created by the state in ultimately unsuccessful attempts at centralization. In his 1989 book on Vijayanagara, Stein devoted considerable attention to the efforts of various rulers to overcome the tradition of segmentary structures, and he considered Vijayanagara “transitional” (1985) to a more centralized state model. Attempts at centralization included early efforts by Devaraya I (AD 1406–1424) to appoint loyal officers to govern forts and provinces, and Devaraya II’s (AD 1424–1446) development of a permanent standing military, heavily reliant on mercenaries, which was under direct royal command. In the sixteenth century Krishnadevaraya also undertook a series of reforms aimed at limiting the autonomy of local elites. This included attempts to create a loyal bureaucracy of Brahman officials, as well as the creation of a stratum of lesser local chiefs (poligars) and officers, who became particularly important in the empire’s upland core (B. Stein 1989a: 43). For Stein, the most significant innovation of the Tuluva period lay in the expanded importance of Telugu-speaking warrior-leaders or nayakas. These individuals were ceded the rights to territories in southern (particularly Tamil-speaking) provinces in exchange for their recognition of royal sovereignty. Within their territories, the nayakas exercised considerable autonomy. As non-local rulers, they undercut traditional practices based on community rights and made new claims on resources that are evident in a number of new tax categories that emerged in this period. Although in principle nayankara rights were not hereditary and could be withdrawn by the king, such withdrawal was difficult to effect in practice. While Karashima


has interpreted the nayakas as evidence for a feudal order in Vijayanagara, Stein argued that the feudal model is not appropriate to South India, as too many additional competing interests cross-cut the administrative segments to allow for a true feudalism. These interests included the important role of temples and the high level of commodity production and exchange. Both the military innovations of the period and processes of urbanization and economic expansion created the possibility for a wide variety of local elites to increase their autonomy and challenge any attempts by the center to limit them. The nayakas may have served as a short-term response to limiting the autonomy of the contentious and wealthy southern reaches of the empire. However, for Stein, in the long term they failed to bring about any effective centralization, and in fact had just the opposite effect. Individual nayakas were able to amass substantial economic and military resources and resisted imperial authority at times of weakness at the center. This was particularly the case after the defeat to the sultanates in AD 1565, when several large and many smaller nayaka polities (Dirks’ “little kingdoms,” 1987) arose. The segments may have been different than earlier ones, but Vijayanagara nonetheless dissolved into its many component parts. Vijayanagara as empire disappeared and was replaced by Vijayanagara as memory and source of legitimacy (e.g., Wagoner 1993). In Stein’s framework, the Vijayanagara center was weak because it could not control its diverse components and because the ruler shared his sovereignty with so many others. Lordship was fragmented, and royal authority was consequently widely dispersed. The Vijayanagara king was in a significant sense the first among equals, who depended upon the ritual recognition of his followers for his authority. By relying upon various subsidiary elites, the empire bore within it the seeds of its destruction. In an interesting echo of some of the early discourses on the Asian state (see chapter 3), Stein contrasted Vijayanagara with its contemporary states in Western Europe, which were, he suggested, “absolutist, bureaucratic regimes” (B. Stein 1989a: xii), though he did not address which fifteenth- or sixteenth-century European states fit this description, how they were organized, or at what scale (it would, in fact, be difficult to identify a European state of the period that might fit his description; Rubiés 2000). There is thus an implicit opposition in his work, between the segmentary South Indian state and the absolutist bureaucratic territorial state that existed elsewhere (see also Perlin 1985: 119). However, in marked contrast to the Oriental despotism models discussed in chapter 3 and their assumptions about the differences between west and east, for Stein the Indian segmentary state is

113

114


characterized by the ruler’s lack of power and his inability to control his followers. While Stein discussed the broad economic trends of the Vijayanagara period, he did not write explicitly of the relations between political and economic structures. These broad trends included the expansion in production and trade in textiles and other commodities, monetization, and urban expansion. Such processes, however, were not the result of state involvement in economic activities, since the state was weak and its power dispersed among many segments. The mechanisms for economic growth are left largely undefined in his discussion, though their consequences – the generation of new sources of wealth with which various elites could engage in political competition – are essential to his understandings of Vijayanagara political structures.

Noburu Karashima: Feudalism Like Nilakanta Sastri and Burton Stein, Noburu Karashima came to the study of Vijayanagara through prior research on the Chola empire. And, like them, throughout Karashima’s writings, Vijayanagara is viewed in contrast to Chola social and political organization. In other respects, however, Karashima’s approach to South Indian historical scholarship has been quite different than both Stein’s and Nilakanta Sastri’s. The latter both took “big picture” perspectives, synthesizing and presenting information from huge areas and time frames. Karashima’s work, on the other hand, is detail oriented; he characteristically dissects a small subset of Tamil inscriptions from a specific time and place. Nonetheless, using these fine-scale inscriptional studies, Karashima has drawn some general conclusions about Vijayanagara rule in Tamil Nadu. His perspective constitutes the third major contemporary theoretical approach to Vijayanagara and can perhaps best be described as “cautiously feudal.” Thus, while Karashima affirms that much more systematic work is necessary before firm conclusions can be drawn about Vijayanagara political structures, he believes that data from Tamil Nadu provide evidence for marked changes in political and social organization during the Saluva and Tuluva periods. These changes correlate with the increased importance of the nayakas, leading towards “the development of a feudal social formation” (Karashima 1992: 8). He does not seek to extend these interpretations to other areas of the empire; rather, his focus is exclusively Tamil. Karashima is ultimately interested in exploring long-term processes of sociopolitical change in later medieval–early modern South India. Like many


Marxist scholars of India, he sought to replace the problematic Asiatic Mode of Production construct with one that allowed for the possibility of significant historical structural change. He has argued that South Indian social formations at the beginnings of the Chola empire are best understood as “state slavery,” in which peasants were controlled directly by the state (a very different view than Nilikanta Sastri’s). In the Middle Chola period we see the beginnings of private landholdings and by the late Chola period, the slavery system has broken down as private land ownership increased further (due primarily to expanded land grants to Brahmans), creating the possibility for an independent peasantry. The state remained dominant throughout the Chola period, but some fissures in the edifice of state slavery appeared. Karashima (1992: 16) argues that the early Vijayanagara period was marked by the continuation of strong central government control in Tamil Nadu, under Vijayanagara governors and generals. These administrators continued and increased the exploitation of the populace, through onerous tax burdens and demands on resources and labor. In response, the inscriptional record provides evidence for peasant resistance, including mass migrations and, in some areas, open revolts (Karashima 1992: 141–158; see chapter 7, this volume). This unstable situation began to change in the late fifteenth century, as the Vijayanagara state replaced these administrators with a new kind of official, the nayaka. For Karashima, the nayakas are best understood as feudal officers, granted rights to land – the nayakkattanam – by the king, in exchange for their commitment to provide military services. Nayakas could themselves be feudatory lords, able to cede territories to “sub-nayakas,” who acknowledged both the Vijayanagara king and the nayaka as their lord. Karashima sees the nayakas as wholly dependent upon the king for their positions; their loyalty to the ruler was expressed in inscriptions recording donations made “for the merit of the king.” This relationship persisted until the late sixteenth century, when nayakas became increasingly autonomous as Vijayanagara central authority weakened after 1565. For Karashima, the appearance of nayaka rule in Tamil regions is both evidence for and part of the process of the strengthening of Vijayanagara power and authority. The imposition of nayakas overcame the problems of the first century of Vijayanagara rule in the Tamil region by creating a class of overlords who profited from investing in the regions they ruled rather than from looting or extracting resources that could not be replenished. As nayakas gained from economic production and increased output in their territories, they took increasing roles in encouraging craft and agricultural production. The earlier pattern of resource extraction by administrators changed to

115

116


investment by holders of territories. As a result, tax burdens were reduced and rebellions and discontent declined (Karashima 1992: 48–49). This was not a feudal order in the sense that it existed in medieval Europe because of the relatively high scale of commodity production and commerce that characterized South India, but for Karashima, the system was comparable to the feudalism of Tokagawa Japan. For Karashima, then, the later Vijayanagara period marked an historic progression from a sociopolitical system of state slavery to one of feudalism. But this was a feudalism marked by a dynamic political economy, in which feudal lords and the state had an important role.

DISCUSSION Through their writings, Nilakanta Sastri, Stein, and Karashima have contributed to scholarly understandings of Vijayanagara and, more generally, of approaches to the South Asian state. They differed significantly in their interpretations of Vijayanagara – as war state, segmentary state, or feudal state, respectively. Yet none of these authors are simple, or simplistic, writers or thinkers, and all acknowledge the complexity of Vijayanagara social, political, and economic structures. These three perspectives also share a predominantly Tamil focus in their approach to Vijayanagara. Stein changed this somewhat in his 1989 book when he suggested connections between Vijayanagara and earlier upland polities, particularly the Hoysalas. But for Stein, as for the others, the Cholas were the template against which to compare Vijayanagara, and the Tamil regions of the empire, areas of substantial wealth in agricultural and commodity production, were a major focus. Nilakanta Sastri saw the Cholas as a highly centralized state, with an efficient, well-organized administration coupled with local community autonomy; this order was necessarily sacrificed to the military challenges of the Vijayanagara period, as power became decentralized with increasing state dependence on subsidiary military chieftains, and local systems of governance were disrupted. Stein saw the Cholas as a more absolute form of segmentary state, with more pervasive ritual kingship than Vijayanagara, where military concerns and the broader economic trends of the period led to increased regional intervention and greater efforts at centralization, efforts that ultimately failed due to the inability to overcome the constraints of the segmentary system. All three authors saw evidence for marked changes in imperial organization with the increasing prominence of nayakas in the southeast in the late fifteenth and


sixteenth centuries, though they disagree on the nature of these changes, and the extent to which it marked a radical break in state structures. A dominant theoretical perspective concerning the nature of the Indian state clearly influenced all of these authors’ writings on Vijayanagara. Nonetheless, each has, to varying extent, acknowledged the many-layered complexity of Vijayanagara-period South Indian polity and society, and the many actors involved in political and economic structures and practices: including kings, local elites, temples, military leaders, merchants, and castes. As noted earlier, only Karashima (1992) explicitly discussed craft producers, in several studies of artisan protests and the granting of various temple and other privileges to artisans (see chapter 7). And while all three authors refer to craft products such as elaborate textiles and architecture, none explicitly address the material dimensions of Vijayanagara political authority. Nonetheless, Stein, Karashima, and to a lesser extent, Nilakanta Sastri, all acknowledge the larger-scale economic processes that characterize the period. For Stein, economic expansion lay largely outside of the political realm, though it created possibilities for state building as well as for the expanded powers of local elites. Karashima also acknowledges economic processes, though he places state administrators, particularly the nayakas, closer to the center of these developments, through their actions to encourage commodity production. However, he also recognizes the role of temples in this realm, and the ability of artisans, merchants, and peasant to act independently of institutions. In their appreciation of the complexity of Vijayanagara-period South India, we see a connection to much recent archaeological and historical scholarship on other imperial states (e.g., Alcock et al. 2001; D’Altroy 1992; Sinopoli 1994a, 2001a). These include the current emphasis on examining the diverse sources of power and authority within imperial systems, and, for many, a marked unease with typologizing. As imperial states incorporate territories through conquest or other means, they encompass areas with distinctive political, social, economic, and ideological histories and structures. In most cases, rather than seek to radically transform those areas to meet a single imperial model, imperial centers accommodate to pre-existing structures so long as security needs and economic demands are met. It is thus futile to attempt to capture imperial structures in a single phrase, such as slave state, or segmentary or patrimonial state, and change over time is to be expected. If a defining characteristic of empires is their messiness, it is no surprise that Vijayanagara also is difficult to classify, or that any attempt to do so, such as the ones discussed above, inevitably raise as many questions as they do answers. Nonetheless, these three perspectives on Vijayanagara

117

118


have framed most historical and, to a lesser extent, archaeological research on the political and economic structures of the Vijayanagara empire, which provide the foundations for my consideration of the political economy of craft production and craft products. Notes 1 I will also use the modern state names, Karnataka, Andhra Pradesh, and Tamil Nadu, to refer to regions inhabited by Kannada-, Telugu-, and Tamil-speakers respectively; these are areas of important ecological, cultural, and historical significance by this period. 2 Not all sources agree on the precise dates of the dynasties or of individual rulers; here I use the chronology presented in Michell 1995.

5

Vijayanagara: sources of evidence

A variety of written and material sources provide evidence for the study of South Indian history, including issues of economy and craft production. Prominent among the written sources are the numerous inscriptions carved on the walls of temples, stelae, and boulders and inscribed on to copper plates. Vijayanagara-period inscriptions were composed primarily in Kannada, Telugu, Tamil, and Sanskrit. Other written sources of the period include diverse South Indian literary works – poetry, temple texts, mythic recountings, and a regrettably small number of administrative documents. We also have written accounts from the northern sultanates and from foreign travelers to South India, including merchants and ministers who visited the Vijayanagara court. Archaeological remains from fourteenth- to seventeenth-century South India are both abundant and remarkably understudied. We know a great deal about the monumental architecture of the period, particularly temples (see especially Michell 1995). Beyond this, however, systematic archaeological research has been conducted only at the first Vijayanagara capital and in its immediate hinterland, and that only since the 1980s. This work has provided scholars with a detailed understanding of the history and plan of the capital, and of the range of economic and other activities and features in its c. 450-square kilometers fortified hinterland or metropolitan region. In this chapter, I address the various sources of written and archaeological evidence available for the study of Vijayanagara.

W R I T T E N S O U RC E S O N V I JAYA NAG A R A H I S TO RY Inscriptions Tens of thousands of inscriptions date to the Vijayanagara period. References to craft production or producers are relatively rare in inscriptions, which primarily record transactions in land, cash, or goods. Based on an examination of more than 3,000 published inscriptions, I have compiled

120


a database of 236 inscriptions relevant to the study of craft production. I describe the database below, after presenting a general overview of the South Indian inscriptional record. Most Vijayanagara-period inscriptions occur in temple contexts, and record donations made to those institutions. Records of temple donations were inscribed on enclosure and temple walls or were engraved on copper tablets that were maintained in temple storerooms. Inscriptional records were also maintained by individual households and lineages, subcastes, or occupational communities; such tablets typically documented property ownership or transactions, and lineage, caste, or cult histories. Smaller numbers of lithic inscriptions are also known from other contexts, inscribed on boulders, gates, and columns near gates and bastions in the city core on reservoir embankments, and on boulders and columns in agricultural fields. The latter have been primarily identified in the Vijayanagara urban core and metropolitan region as the result of the intensive archaeological work that has occurred there since the late 1970s (Nagaraja Rao and Patil 1985; Patil and Patil 1995, 1997). These inscriptions record donations, place or feature names, field boundaries, reservoir construction, and, occasionally, the names of individuals (i.e., graffiti). Similarly intensive survey work in other regions would no doubt reveal many more inscriptions, but such research remains to be done in much of South India. As a result, at present, the vast majority of documented inscriptions come from temple contexts, and, not surprisingly, the role of temples in fourteenth- through seventeenthcentury South India has received considerable scholarly attention. Given the biases of the inscriptional record, it is likely that the importance of temples, while clearly significant, has been overemphasized in traditional historiography (see Narayana Rao, Shulman, and Subrahmanyam 1992; Morrison and Lycett 1997), while the role of institutions and individuals whose records were of less durable form has been underestimated. Since inscriptions comprise the most studied source of historical evidence for the Vijayanagara period, it is critical that this bias be acknowledged, particularly when assessing the overall role of temples in South Indian political economy and society. Vijayanagara-period inscriptions were recorded in at least four languages – Sanskrit, Kannada, Telugu, and Tamil. In the case of the three Dravidian languages, the language used was typically the dominant spoken language of the region where the inscription was carved. There was thus not a single language of temple administration. When Sanskrit inscriptions occur, they are sometimes duplicates of texts in a local language. In some cases, more than one language is used in a single inscription – often Sanskrit in the formulaic opening passage and the local spoken language to


describe a specific transaction. In other cases, one language may be written in the script of another (e.g., a Kannada text written in Telugu script; Morrison and Lycett 1997). While, as mentioned, records of temple donations predominate, inscriptions occasionally record other kinds of information – concerning conflict resolution, economic agreements, tax remissions, or the conferral of titles or privileges. Most inscriptions, of all categories, contain information on date, local or imperial rulers, and place names, along with formulaic opening and closing passages and precise details on the transactions being documented. Occasionally, witnesses to the transactions are named, and less often, inscriptions include the name of the artisan who composed and/or inscribed the text. The documentation of Indian inscriptions began in the late 1800s and is still ongoing. Compendia of recorded inscriptions have been published for more than a century. Major series include South Indian Inscriptions (SII, 1890–1992), Epigraphia Indica (EI, 1888–present), and Annual Reports on Indian Epigraphy (ARIE, 1887–); regional series include Epigraphia Carnatica (EC, 1886–present), Epigraphia Andhrica (1969–), and Inscriptions of Andhra Pradesh (1966–). Other relevant publications include recent volumes of Vijayanagara inscriptions produced by the Karnataka Department of Archaeology and Museums (Vijayanagara Inscriptions, 1985, ed. Gopal; Inscriptions at Vijayanagara, 1995, and Inscriptions of Bellary District, 1997, ed. Patil and Patil), and corpuses of inscriptions from major temple centers, such as Tirupati-Tirumala (Tirumalai-Tirupati Devasthanam Epigraphical Series (TTDES) 1931–38), and Tiruvannamalai (Srinivasin 1990), and the Vitthala temple at Vijayanagara (Filliozat and Filliozat 1988), among others. While these sources are invaluable, their very variety is also problematic. Inscriptions have been published in a wide variety of formats. Most often, complete transcriptions of texts are prepared, typically with brief English language summaries. Full translations are less common, and some publications provide only brief summaries in English and local languages. In many cases, the same inscriptions appear in multiple sources; typically, their publication lineage is traced, and they are referenced by their first appearance in print. Because of the many sources in which Vijayanagara-period inscriptions have appeared, it is difficult to come up with a precise number of published inscriptions. Trautmann et al. (1985) have estimated that there are between 80 and 100,000 published inscriptions from South India. A significant percentage of these date from the fourteenth through seventeenth centuries. To date, the most valuable analyses of inscriptional sources are those that have focused on relatively small subsets of inscriptions from specific locales and

121

122


periods (see particularly, the work of Karashima 1984, 1992). However, several scholars are attempting to develop systematic computerized databases of South Indian inscriptions (e.g., Trautmann et al. 1985 and Orr 2000 and Heitzman 1997 for the Chola period; Talbot 2001 for the Kakatiyas, and Morrison and Lycett 1994, 1997 for the northern regions of the Vijayanagara empire).

The inscriptional database My database of 236 craft-related inscriptions (Figure 5.1) is based on the examination of 3,184 inscriptions from several publications. These include the Tirumalai-Tirupati Devasthanam Epigraphical Series (TTDES 1931–38, with approximately 1,100 Vijayanagara-period inscriptions), the Tiruvannamalai inscription volume (Srinivasan and Reineche 1990; approximately 150 Vijayanagara-period inscriptions), the 17 volumes of Epigraphia Carnatica (EC, New Series, with 1332 Vijayanagara-period inscriptions), and the Inscriptions of Bellary District (Patil and Patil 1997, 171 Vijayanagaraperiod inscriptions) and Inscriptions at Vijayanagara (Patil and Patil 1995, 471 Vijayanagara-period inscriptions). With the exception of the last two works, which provide only English summaries, all volumes examined provide full English translations of the inscriptions reported. I also include an additional 38 inscriptions specifically relating to craft production or producers that have been published by Ramaswamy (1985a, 1985b) and Karashima (1992); these are predominantly Tamil inscriptions, which are otherwise under-represented in my sample. This database is by no means a complete survey of all published Vijayanagara-period inscriptions, nor even of all inscriptions in which artisans are mentioned. However, it does encompass the full corpus of published inscriptions from several areas of the empire, including the imperial capital and its hinterland and three major temple centers in the southeastern portion of the empire, an area that experienced significant intensification of craft production activities. The most bounded and complete sets of inscriptions in the database come from complete epigraphic surveys of the temple complexes at Tirupati, Tirumala, and Tiruvannamalai. The temple complexes in the town of Tirupati and the nearby Tirumala hill located in southeastern Andhra Pradesh are today among the largest and wealthiest in South Asia. Their rise to prominence began during the Vijayanagara period, when the Venkateshvara temple at Tirumala was patronized by a number of Vijayanagara rulers. From 1922 to 1927, the temple administration documented the lithic


123

ANDHRA PRADESH

KARNATAKA

Raichur

Koppal

Kurnool

VIJAYANAGARA

Be

ry lla Anantapur

Cuddapah

Shimoga Tirupati

lep ut

Chittoor Hassan

Tiruvannamalai

Ch

Co or g

ing

North Arcot

Mysore South Arcot

ALA KER

tore Coimba

Tiruchirapalli Thanjavur

TAMIL NADU

0

km

5.1 Areas included in inscriptional database

inscriptions at the temples, and published them in the six-volume TirumalaiTirupati Devasthanan Epigraphical Series (TTDES 1931–38). These volumes contain the complete text of 1,066 Vijayanagara-period inscriptions in original language and full English translation. Of the total, 640 are from the large Sri Venkateshvara temple complex on Tirumala hill, and 340 from the Sri Govindaraja temple in Tirupati; the remainder come from smaller temples in the region (TTDES 1931: 1). More than 3,000 sixteenth-century copper plates are also stored in the temple complex, but these have not been published. The vast majority of the published inscriptions date to

300

124


the Vijayanagara period, with most dating to the sixteenth century. Four hundred and eighty inscriptions date from 1509–1542, spanning the reigns of just two Vijayanagara rulers: Krishnadevaraya (c. 1509–1529, n=229 inscriptions) and Achyutadevaraya (1530–1542, n=251). Of the 1,066 Vijayanagara-period inscriptions, 59 (5.3 percent) make explicit reference to craft production or producers; all are included in the database. Tiruvannamalai is a large Shaivite temple complex located in North Arcot District, Tamil Nadu. Dating from the seventh century AD to the present, it was a recipient of royal and elite patronage throughout the Vijayanagara period. Approximately 500 inscriptions have been recorded, including both lithic and copper plate records; 150 date to the Vijayanagara period (Srinivasan and Reiniche 1990: 81). Only 2 of the 150 inscriptions (1.3 percent) contain direct references to craft producers, in both cases Kaikkolar weavers (Srinivasan and Reiniche 1990: 438, 458). Initial publication of Epigraphia Carnatica (EC) began in 1886 under the supervision of B. Lewis Rice. Volumes were issued until 1934 under the Mysore Archaeological Department (with the final volume edited by M.H. Krishna). In the 1970s, the series was reissued and the original seventeen volumes were condensed into eight. Translations were updated and supplementary sections or volumes were added containing inscriptions that were recorded after the initial volumes was published. The works are organized at the district level and focus on the following districts of southern and central Karnataka: Coorg, Shimoga, Mysore, Sravanabegola, Hassan, and Mandya. They include 1,332 inscriptions dating between 1350 and 1700. Of these, 123 (9.2 percent) include references to craft production or producers. Throughout the 1990s, Dr. C.S. Patil of the Karnataka Directorate of Archaeology and Museums undertook the systematic publication of inscriptions documented at Vijayanagara and surrounding districts. Many of these inscriptions have been newly identified through the work of the Karnataka Department, the Archaeological Survey of India, the Vijayanagara Research Project, and the Vijayanagara Metropolitan Survey, whereas others have been previously published in other works. In 1995, Patil and Patil published the first volume of the Inscriptions of Karnataka Series, which contained the 471 known inscriptions from the site of Vijayanagara, and in 1997 they published Inscriptions of Bellary District, which contains 657 inscription, of which 171 date from AD 1300 to AD 1700. These publications include a brief English summary of the inscriptions, rather than their full text. As noted above, the remaining inscriptions in the database primarily come from northern Tamil Nadu and include a grab bag of inscriptions that


make mention of craft producers from publications by Vijaya Ramaswamy (1979, 1985a, 1985b, 1985c, 1993) and Noburu Karashima (1992). While not a comprehensive review of all published inscriptions, this list most likely contains a significant portion of references to craft producers. The relatively infrequent references to craft producers in the Vijayanagara inscriptional record is presumably a consequence of both the relatively low social and economic status of many producers and their limited participation in temple affairs. For each inscription, the database includes information on location and context (state, district, temple), language, date (to year or century, depending on information available), inscription type, craft groups mentioned, content, and citation. For the purposes of this study, I defined the variable “inscription type” according to how craft producers were referred to in the inscription, rather than the overall intent of the inscription (e.g., to record a donation or other transaction). The inscription types identified and their frequencies in the database are summarized in Table 5.1. The frequencies summarized in Table 5.1 indicate that a very small number of categories dominate the inscriptional references to craft production, with three inscription types – artist’s signature, donation of taxes, and distribution of offerings – providing 164 (69.5 percent) of the 236 inscriptions in the database. The first category records the actions of artisans regarding claims to authorship of their products; the latter two record the actions of elites and temple administrations, and craft producers appear only as people who owe taxes or receive payment for their labors. Artisans are more visible as social actors in the less common inscription types, where they appear as temple donors, direct recipients of tax remissions, or engaged in social protest. I will consider the content and patterns in inscriptional references to craft production in more detail in chapters 6 and 7.

Documents In addition to inscriptions, numerous other written sources provide information on the Vijayanagara period. These include the writings of foreign travelers to South India as well as indigenous poetry, biographies, and sacred texts. Each of these sources poses interpretive challenges. The writings of foreign travelers were affected by their linguistic limitations, incomplete understandings of South Indian political and social organization, individual interests and background, and their desires to please their intended audiences back in Europe (Teltscher 1995; Campbell 1988). The diverse Vijayanagara

125

126


Table 5.1 Inscription types Type

Number

Description

1. Artist’s signature

57

2. Donation of taxes

55

3. Distribution of offering

52

4. Temple donation 5. Tax remission

15 14

6. Social protest 7. Other

13 5

Found on sculptures or inscriptions; includes both engraver and author of text. Refers to donation or reallocation of taxes by temple donor (usually local or imperial administrator) to named temple. Specific taxes are designated that include taxes on craft producers, i.e., taxes on potters or basket makers, or loom taxes. Refers to how proceeds that will be generated from a specific donation are to be distributed. Names specific categories of individuals affiliated with temples who will receive proceeds. For example, payment to Kaikkolar weavers to decorate shrines for specific festivals. Refers to donation made by craft producers. Refers to granting of tax remission to artisans, often as incentive to settle in temple town or in response to social protest. Protests by artisans, often against unfair taxation. References to weights and measures, gifts to artisans, lands of artisans, named places (e.g., “the site of the poets”), etc. Refers to granting of temple privileges to artisans. Such privileges include the right to use ritual paraphernalia or participate in festivals. Refers to grants of land to artisans, often as incentives to settle in particular district or temple town. Refers to establishment of tax rate on specific artisan activities or tools. Often in response to social protest. Refers to requests for temple privileges made by artisans.

8. Grants of temple privileges

5

9. Land to artisans

4

10. Tax rates

3

11. Requests for temple privileges 12. Inter-caste dispute

2 2

13. Payment for labor 14. Provisioning of well

1 1

15. Legal agreement

1

16. Rental agreement 17. Eulogy of artisan 18. Witness to transaction 19. Donation of labor and goods

1 1 1 1

Refers to resolution of conflicts involving artisans, includes detailed description of the dispute and its resolution. Records payment to artisans for specified services. Records community agreement to provision a well for pilgrims and travelers. Refers to agreement among artisans; probably results from inter-caste dispute, but dispute is not mentioned. Records rental agreement of property. Tribute to deceased artisan. Mention of artisans as witnesses to transfers of property. Records donations of labor or goods made by artisans to charity.


Table 5.2 Foreign travelers in fourteenth- through seventeenth-century South India Visitor

Origin

Dates

Translation

Mahuan Nicolo Conti (as told to Poggio Bracciolini) Abdul Razzaq Athanasius Nikitin Duarte Barbosa Domingo Paes Fernao Nuniz John Van Linschoten

China Italian

AD 1409 early 15th c.

Phillips 1896 Major 1857

Samarkand Russen (Twer) Portuguese Portuguese Portuguese Dutch

AD 1442 late 15th c. AD 1506–1518 AD 1520–1522 AD 1535–1557 1580s–1592

Thackston 1989 Major 1857 Dames 1989 (1918) Sewell 1900 Sewell 1900 Burnell 1988

literary sources were primarily courtly or temple texts intended to communicate specific messages to specific audiences. Further, many have been transcribed multiple times, with numerous additions and subtractions. A third useful category of written evidence post-dates the Vijayanagara period. Included here are the writings of late eighteenth- and early nineteenth-century British colonial officials, such as Frances Buchanan, Colin Mackenzie, and Thomas Munro (Stein 1989b), among others, who gathered local texts and wrote extensively about South India and its economic resources. I address these three categories of written evidence in turn.

Travelers’ accounts The Vijayanagara capital and empire were visited by travelers from Europe, Central Asia, and East Asia. By the early sixteenth century, the Portuguese had taken several ports in western India, and had established their control in Goa. As they attempted to gain control of maritime trade in the region, Portuguese officials, merchants, and clergy were involved in a variety of relations with the Vijayanagara court. Several foreign visitors to South India later published accounts of their travels (Table 5.2). For the most part, these accounts contain relatively little information on craft production activities or producers, but they do provide some information on the abundance and diversity of craft products at the imperial capital and throughout the empire. Among the limited references to craft producers in the travel literature is Portuguese merchant Domingo Paes’ mention of a “street where there are many craftsmen, and they sell many things” (in Sewell 1900: 256). He

127

128


further observed, “there are temples on every street, for these appertain to institutions like the confraternities you know of in our part, of all the craftsmen and merchants” (in Sewell 1900: 256). Occupational specialists mentioned by Paes included warriors, wrestlers, musicians, washers, merchants, and boatmen. Paes’ mention of “confraternities” may be a reference to organizations such as the Ayyavole 500, discussed in chapter 4. Fernao Nuniz, a second Portuguese merchant who visited Vijayanagara in the early sixteenth century, also left a detailed written account (see Sewell 1900). His writings were largely concerned with the empire’s political history. However, he also included descriptions of the king’s daily routine and the capital’s military forces. He noted that the army included “two thousand artificers, namely blacksmiths, masons, and carpenters, and washermen who wash cloths” (quoted in Sewell 1900: 381). Nuniz also described women dancers, bearers, wrestlers, astrologers, accountants, judges, bailiffs, guards, musicians, cooks, and attendants, all of whom were employed in the royal palace (Sewell 1900: 382–383). Abdul Razzaq visited Vijayanagara in the early fifteenth century as ambassador from the Timurid court of Samarkand. In describing the city, Razzaq referred to the “craftsmen of the bazaar” (in Thackston 1989: 309) and noted that “the practitioners of every craft have stalls adjacent to one another. The jewelers sell pearls, rubies, emeralds, and diamonds openly in the bazaar” (quoted in Thackston 1989: 308). And in celebrating the workmanship visible in the royal palace, he wrote of the royal throne, “it was immense, of gold studded with precious gems and minutely embellished with artifice. There is no place on the face of the earth where they do gem studding better than there” (Thackston 1989: 314). Razzaq also noted the presence of male and female musicians, storytellers, dancers, and jugglers. Poggio Bracciolini’s recounting of the tales of the early fifteenth-century Italian traveler Nicolo Conti described the mining of gold and semi-precious stones in South India. This was a labor-intensive activity that involved large numbers of laborers who screened and sifted sandy sediments in resource rich areas. He noted that “great care is exercised by the masters to prevent theft by the workmen and servants – overseers being appointed, who not only shake the clothes of the operators, but even examine every part of their persons” (quoted in Major 1857: 30). Unfortunately, he did not record how laborers or their overseers were recruited or compensated, or who received the output of their labor. While providing relatively little detail on craft producers and their activities, several of the foreign accounts of the Vijayanagara capital and empire provide rich information on the range and diversity of goods produced and


available in markets. Frequently mentioned are minted coins, gold ornaments, precious stones (rubies, emeralds, diamonds, pearls), musical instruments, and elaborate woodwork inlaid with precious stones. Paes described some of the military equipment he observed during royal processions including ornate horse and elephant trappings, swords, battle-axes, javelins, iron armor, bows and arrows, muskets, and “blunderbusses” (quoted in Sewell 1900: 275–280). He and other travelers also described the sophisticated stone masonry of Vijayanagara’s fortifications and the ornamented architecture of the royal palaces, decorated with elaborate sculpted or carved plaster columns and numerous wall paintings (e.g., Sewell 1900: 244, 284, 287). Textiles appear prominently in many descriptions of the capital, with the opulence and diversity of cotton, linen, and silk cloths noted by virtually every visitor. Paes described the elaborate textiles that adorned buildings, royal and elite garments of fine cotton ornamented with gold embroidery, and the velvet and damask umbrellas that shielded royal figures. He also noted the presence of imported Chinese silk and Mecca velvet at the capital. The importance of textiles to royal ceremony and courtly life is further emphasized in Paes’ description of the disposition of cloths at the end of the annual Mahanavami festival, when everyone put on “new rich and handsome garments” and “all the captains gave their men cloths of many colors, some having his own color and device” (Sewell 1900: 281–282). Van Linoschoten, who visited the southeast coast of India in the late 1500s, noted the “divers sortes of loome workes and figures, verie fine and cunningly wrought” (Burnell 1988: 91). Vijayanagara’s foreign visitors, like contemporary Indian authors, were far more concerned with activities and commodities associated with the imperial court and urban elites than with non-elite goods or communities. Their accounts provide little information on production activities or craft producers. They do, however, provide evidence of a society characterized by high degrees of occupational specialization and highly structured spatial organization, with an extraordinarily rich and diverse range of material culture. Paes’ mention of streets of artisans at the imperial capital suggests that at least some craft goods were produced within Vijayanagara’s city walls.

Literary works Numerous literary works can be traced to the Vijayanagara period. Written in Telugu, Kannada, Tamil, and Sanskrit, many of these documents have been transcribed numerous times since their initial composition and exist as palm

129

130


leaf and paper manuscripts in libraries and private collections throughout South India. Several have been published in their original language or in English translation, and a number of Kannada manuscripts have been recently summarized in English by Dr. C.T.M. Kotraiah (Kotraiah n.d.a–n.d.l; unpublished manuscripts in possession of author). The extant Vijayanagara-period manuscripts were rarely formal economic or historical texts, though Stein (1989a: 86–91) has suggested that census data or economic records (e.g., numbers of looms or farmers per community) were maintained in provincial centers and possibly at the imperial capital for taxation purposes. Such texts would be of tremendous value, but at present, only one – the Alamkonda Puttasti, discussed in chapter 4 – is known (see also pp. 258–262). Instead, most preserved texts are religious treatises, romantic or sacred stories and songs (often retellings of ancient Sanskrit tales), and royal biographies written in prose and verse. Their authors included court poets (attached to kings, high-status royal officers, or provincial leaders, see pp. 160–166), temple poets, religious devotees, and male and female elites. Among the latter are works attributed to Gangadevi, a fourteenth-century Vijayanagara queen (the Mahdhuravijaya), and to some of Vijayanagara’s kings, including Devaraya II (AD 1424–1446, the Sobagina Sone) and Krishnadevaraya (AD 1509–1529, e.g., the Amuktamalyada and Jambavati Parinayam). Authors of literary works are among the few specialists whose names were recorded during Vijayanagara times. The names of other producers of fine arts, such as sculptors or painters, are less often preserved. Many Vijayanagara literary works contain valuable information on urban life and royal activities. Some refer directly to Vijayanagara or other contemporary cities. Others are set in mythic or historic cities, but likely nonetheless reflect their authors’ experiences or idealizations of contemporary urban spaces and social and economic relations. Various specialists are mentioned in many of the texts listed in Table 5.3. Since the majority of literary works refer in one way or another to courtly life and activities such as royal weddings or royal audiences, it is not surprising that the lists of specialists are both limited and highly redundant. The redundancy among various texts lends support to the assumption that they are, at least in some sense, reporting on the (ideal) makeup of royal courts, military camps, and other royal settings. Diverse entertainers figure prominently in many texts, including poets, bards, composers, male and female musicians, female dancers and attendants, wrestlers, and theatrical performers. Others frequently mentioned as court attendees include subsidiary rulers and their entourages, military


131

Table 5.3 Selected Vijayanagara-period literary works Title

Author/Date

Nature of Work

Padmanaka, lived in Ambapura (unknown), descendent of Hoysala rulers, c. 1385 AD

Story of author’s ancestor who was a minister in court of Hoysala king (Kotraiah n.d.f).

Sobagina Sone

Devaraya II, Vijayanagara emperor, AD 1410

Romantic stories, as told by poet to his wife (Kotraiah n.d.k).

Pamasthana Varnanam

Chandrasekhara (Chrakavi), court poet of Devaraya II, AD 1430

Description of Pamapara (Virupaksha temple and associated settlement area at Vijayanagara capital; Kotraiah n.d.g).

Airavat

Kumara Vyasa, Smartha Brahman, resident of Gadag, belonged to family of village official, c. AD 1430

Recounting of an episode from the Mahabharata, story of Airavata, the celestial elephant of Indra (Kotraiah n.d.a).

Sivatatva Cintamini

Lakkandandesa, minister and provincial governor, military leader in the court of Devaraya II, c. AD 1450

Historic poetic work about Shiva and the Kalachuri dynasty, numerous references to Vijayanagara city and suburbs (Kotraiah n.d.j).

Nemijinesa Sangata

Mangarasa, subordinate ruler/governor of Kallahalli, AD 1508

Life of a Jaina saint (Kotraiah n.d.e).

Samyukta Koumudi

Mangarasa, subordinate ruler/governor, c. 1509 AD

18 short stories about religious concerns and moral life (Kotraiah n.d.i).

Ramanantha Charite

Nanjunda Kava, son of feudatory ruler, c. AD 1525

Story of Ramanatha, prince of Kampili kingdom, hero in battles against Delhi sultanate (Kotraiah n.d.h).

Mohanatarangini

Kanakadasa, minister-governor (see Romantic poetic work, retelling of life of Songs of Kanakadasa, below) under Krishna, dedicated to Vijayanagara ruler Vijayanagara king, song writer and Krisnadevaraya (Kotraiah n.d.d). poet, Dharwad District, of hunter’s caste, AD 1550

Vijayakumari Charite

Srutakirti, Jaina ascetic, Mysore District, AD 1567–1568

∗

I. Kannada Padmaraja Purana

Biographic poem of Hoysala woman named Vijayakumari, retelling in Kannada of original Sanskrit texts (Kotraiah n.d.l).

Channabasava Purana Virupaksha Pandita, head of Virupaksha Temple (Vijayanagara), AD 1585

Religious work (Kotraiah n.d.b).

Kanthirava Narasrja Vijaya

Biographic sketch of one of rulers of Mysore Odeyars (Kotraiah n.d.c).

Govinda Vidya, court poet, Srirangapattana, AD 1648

(cont.)

132


Table 5.3 (cont.) Title

Author/Date

Songs of Purandaradasa

Purandaradasa (AD 1485–1565), Vaishnava devotional songs, about 1,300 resided at Vijayanagara and Tirupati known (W. Jackson 1998).

Songs of Kanakadasa

Vaishnava devotional songs (W. Jackson Kanakadasa (1500s, said to have lived 98 years), visited Vijayanagara, 1998). minister-governor under Vijayanagara king, song writer and poet (see Mohanatarangini, above), Dharwad District, of hunter’s caste, AD 1550

II. Telugu Rayavacakamu

Nature of Work

author unknown, probably resided in the court of Madurai nayaka kingdom, c. 1595–1602

Purports to be diplomatic report from the court of Krishnadevaraya, though in fact composed c. 70 years after his reign (Wagoner 1993).

Amuktamalyada

Attributed to Krishnadevaraya (AD 1509–1529)

Document on kingship reportedly composed by Krishnadevaraya, only small portions have been translated and published (Nilakanta Sastri and Venkataramanayya 1939).

Songs of Annamacharya

“master of devotional songs” at Tirupati, AD 1408– , composed more than 30,000 songs

Vaishnava devotional songs (W. Jackson 1998)

The Kaliahistisvara Stakamu of Dhurjati

Temple poet of Kalahasti, 16th century

Collection of short poems dedicated to temple god (Heiftiz and Narayana Rao 1987).

Gangadevi, 14th-century queen of king Kampana

Biographical poem descriping palace life (Sridhara Babu 1975; Hamper Hiebert 1985).

Varadambikaparinaya (The Marriage of Varadamibka)

Tirmalamaba, 16th-century queen, wife of Acyutaraya

Biographical poem, description of royal court of Acyutaraya (Hamper Hiebert 1985).

Ragunatha-bhyudaya (The Feats of Ragunatha)

Ramabhadrama, 17th-century queen, wife of Ragunatha

Biographical poem, description of royal court of Ragunatha (Hamper Hiebert 1985).

III. Sanskrit Madhuravijaya

∗

Except where noted, sources consulted for these published and unpublished Kannada works are unpublished English summaries recently compiled by Dr. K.T.M. Kotraiah for the Vijayanagara Research Project, directed by John M. Fritz.


leaders, ambassadors, accountants, scholars, astrologers, concubines, doorkeepers, servants, and guards. Only one document mentions a producer of material crafts as part of a courtly assembly. The Ramanatha Charite (AD 1525) describes a goldsmith as a regular participant in royal ceremonies in the Vijayanagara court. Goldsmiths often worked as moneylenders during the Vijayanagara period and it is not clear in what capacity this individual was serving in the court. Urban centers are consistently described in the texts as heavily fortified and enclosed by stone masonry walls with numerous ramparts and watchtowers, and as surrounded by gardens and agricultural fields. Particular attention is paid to water sources, particularly reservoirs, which are described as beautiful and bucolic locations, and as trysting or sporting places for royal men and women and their consorts. Cities are portrayed as beautiful and well laid out, with elegant and luxurious houses adorned with flags, banners, flowers, and jewels; murals also are frequently reported to decorate the walls of houses, palaces, and temples. The texts frequently depict cities as organized into specialized areas with streets of shops and artisans. Among the goods mentioned as sold in shops are precious stones, including rubies, diamonds, emeralds, and pearls, clothes of silk and cotton, gold ornaments, bronze and brass objects, aromatics and perfumes (sold by women merchants, according to Padmananka’s Padmaraja Purana; Kotraiah n.d.f), and flowers. Several authors mention streets of the Sun and Moon along which artisan workshops and stores were located. An especially valuable work describing urban layout refers not to Vijayanagara but to the later capital of Srirangapattana, near Mysore. This work, the Kannada Kanthirava Narasrja Vijaya, was composed by the court poet Govinda Vidya for his patron Nanjarajendra (described as the ministergeneral of the army) in 1648 and contains vivid descriptions of Srirangapattana (Kotraiah n.d.c). References are made to the diverse inhabitants of the city: “citizens of different traditional classes, but [each residing] in their respective streets” (Book II, verses 72–74). Among the groups listed as residing within the city walls are perfume vendors, bronze smiths, oil millers, barbers, coppersmiths, fishers, painters, silk cloth weavers, silk sari weavers (of paithani tradition), instrumentalists, vocalists, dance masters, bead makers and/or sellers, masons/brick layers, torch bearers, various classes of warriors, and concubines (Book IV, verses 54–63). A biographical poem by the Jaina ascetic Srutakirti (AD 1567; Kotraiah n.d.l) discusses a city (believed by Kotraiah to be Vijayanagara) containing shops of a range of merchants including goldsmiths, bronze merchants, textile merchants, and vendors of herbal medicines, sandalwood paste, and

133

134


flowers. The text notes that “areas outside the fort wall were inhabited by people of the lower classes such as washermen, potters, barbers, carpenters, and others . . . subjects of all the above castes live in streets reserved for them as per the established customs of the society and without giving room for any violation of them” (Book I, verses 103–106). The Channabasava Purana is a religious work authored by Virupaksha Pandita, the head of the large Virupaksha temple center at Vijayanagara and dated to AD 1585 (after the city’s abandonment; Kotraiah n.d.b). It includes descriptions of the Vijayanagara city, markets, and military forces. Mention is made of the royal palace built by “qualified architects and as per the architectural principles contained in the treatises” (Book XI, verse 48), and of various royal attendants including bards, royal eulogists, and gardeners. This work also provides information on military encampments, including discussions of specialization within military forces and among the components of the military encampment. These camps are described as containing traders, goldsmiths, cloth merchants, jewelers, cooks, flower merchants, tailors, cobblers, sharpeners of weapons, and oil merchants, among others. Blacksmiths are reported to have accompanied armies in order to manufacture and repair a range of weapons. Streets of hunters are mentioned in Devaraya II’s Sobagina Sone (c. AD 1410; Kotraiah n.d.k) and professional hunters (betaya bedaru) are described in the Ramanatha Charite of Nanjunda Kava (c. AD 1525; Kotraiah n.d.h). The hunters are described as stout with thick reddish moustaches and as guiding the king on royal hunting expeditions. They lived within the city and were engaged in such tasks as manufacturing nets, repairing bows and arrows, and selling live and dead animals. Artists and engravers are mentioned in the Mohanatarangini of Kanakadasa (AD 1550, Kotraiah n.d.d; Book III, verses 34–41), including male and female painters, diamond cutters, and sculptors of bronze and stone, who sold their products in market stalls. The text also refers to portraits, murals, and paintings on banners and flags. The Airavat by Kumara Vasa (AD 1430; Kotraiah n.d.a) recounts the story of Indra’s sacred elephant. The worship of a clay figurine of an elephant is recounted; this image is described as formed of black clay, and then painted by professional painters and ornamented with gold. Along with references to occupational groups, literary works typically describe a wide range of craft goods. These are often associated with royal courts and include many kinds of elaborate ornaments and garments, along with numerous musical instruments, gold and silver serving vessels, and so on. Many types of weapons are also mentioned.


The sample of literary works discussed here comprises only a very small portion of the writings that can be attributed to the Vijayanagara period. Although missing many of the details we might wish for, these texts provide useful information on the highly developed occupational specialization that characterized South India during this period. Spatial segregation of occupational groups in urban residential districts, as also reported by Paes, is well documented. Some information is also provided on the gendered division of labor, with mention of women artists, dancers, musicians, palace guards and attendants, and hunters. As noted above, references are primarily to elite crafts and artisans and little information is provided on productive organization or producers of non-elite goods. Nonetheless, the texts do provide a vivid sense of the socioeconomic complexity and the vibrancy of urban and courtly life during the Vijayanagara period.

Colonial accounts The last category of texts I consider postdates the Vijayanagara period but nonetheless contains valuable information on technologies and distributions of craft producers in South India. These are texts by various British colonial functionaries who traveled throughout South India in the late eighteenth and early nineteenth centuries. Particularly important for the study of craft production is the three-volume compilation by Francis Buchanan first published in 1807 (reprinted 1988) and entitled (in full): A Journey from Madras through the Countries of Mysore, Canara, and Malabar Performed Under the Orders of the Most Noble the Marquis Wellesley, Governor General of India, for the Express Purpose of Investigating the State of Agriculture, Arts, and Commerce; the Religion, Manner, and Customs, the History Natural and Civil, and Antiquities, in the Dominions of the Rajah of Mysore and the Countries Acquired by the Honourable East India Company, in the Late and Former Wars, from Tippoo Sultain.

In keeping with the East India Company’s economic interests, Buchanan began his journey in 1800 with instructions to focus particularly on agricultural resources and technologies, minerals and mining, and “manufactures and manufacturers.” His three-volume work is organized geographically, recounting the route of his travels. The volumes contain valuable descriptions of mining and smelting technologies, production of building plaster or quicklime or cancar, stone quarrying, glass production, weaving and dyeing, and tanning. By the time of Buchanan’s travels, British “collectors” were well established in many parts of the South. Regional trade routes focused

135

136


either on Srirangapattana, the former capital of Tipu Sultan near Mysore, or were linked to the British East India trade. It is difficult to compare the scale or intensity of production of various goods at the start of the nineteenth century to that of the Vijayanagara period, and certainly textile manufacture and other productive activities tied to the international economies had changed dramatically. Nonetheless, the information on technologies such as iron working (Figure 5.2) or lime production conforms well with the archaeological evidence from the Vijayanagara metropolitan region, and information on the distribution of weaving communities can be compared to Vijayanagara inscriptional records and other sources. I will present details from Buchanan’s writings in discussions of individual technologies in chapter 6. In the third volume of his work, Buchanan reproduced a census carried out by a Mr. Ravenshaw, supervising officer or “collector” of the ten taluks (geographic divisions within districts) of the southern part of Canara province (modern Karnataka). The census provides summary information on 396,672 inhabitants of the region by “caste or trade,” age (adult vs. child), and sex. One hundred and twenty distinct caste or occupational groups are described, including several categories of agricultural laborers, Brahmans, and a variety of artisans. Of the latter, potters are the most numerous. I reproduce a small subset of the census report in Table 5.4, focusing primarily on various craft producers. Much has been written about the problematic nature of censuses and gazetteers in colonial India and their role in what has been referred to as the “objectification of India” (of which Buchanan’s work is a prime exemplar; see for example, Cohn 1996; Appadurai 1986; Pels 1997). Through such surveys, knowledge was gathered on a range of subjects deemed important by the British. This information was then “transformed into usable forms, such as published reports, statistical returns, histories, gazetteers, legal codes, and encyclopedias” (Cohn 1996: 5). Cohn and others have argued that the colonial project of counting and categorizing led to the formalization and reification of social categories and a particular construction of India that served British colonial interests. The problems of such sources are indisputable; however, they also indisputably captured valuable information on recognized social and occupational distinctions. Here I simply present information from one to present a portion of the range of occupations recorded, and their relative numbers in one region of South India in the early 1800s. The extent to which people labored in their named occupations and/or their identities were fluid (or misrecorded) can not be addressed.


5.2 Iron smelting and forging facilities depicted by Frances Buchanan (1988[1807], vol. I: 170)

137

138


Table 5.4 Summary of South Canara census data, 1807 (craft producers in bold) No. 1 3 14 15 31 89

96 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122

Caste or trade Brahmans Kankanes Bankers, shopkeepers and traders Jainas Cultivators Bunts Culitvators Biluaras People who extract the juice of palms Mistries Head carpenters (term can also refer to builders or masons) Dererd (Whalliaru) Slaves employed in cultivation Hujam Barbers Chummar Workers in leather Sungtrash Stone-cutters Sunar Gold and silversmiths Cassar Workers in brass Lohar Blacksmiths Julai Weavers Canara Kumbhara Pot-makers Kuddai Carpenters Rungara Dyers Boradur Mat-makers Tambutgars Coppersmiths Chitragar Painters Pujar Cotton cleaners Shiculdars Cutlers Zeendar Saddlers Dirzi Tailors Toipha Dancers and musicians Jetty Wrestlers Killabund Architects, literally constructors of forts Tapegar Jewelers Jilligar People who search wells and tanks for lost money Moothaley Adagathur Mogayar Boatmen and fishermen Corchey Day laborers

houses

men

boys

women

girls

total

6807 2434

12677 4724

6932 2419

13192 4495

4080 1436

36881 13074

2700 8183 11397

5108 19348 20222

2307 7775 8087

4763 19041 19376

1914 6654 6079

14092 52819 53764

14

26

13

23

4

66

12278

16751

7528

16633

6446

47358

517 193 27 1329 127 127 847 2188 602 1 65 5 5 16 10 32 125 156 2 4

912 386 48 2714 234 210 1367 3892 986 4 111 13 9 27 26 62 252 140 5 3

352 187 16 1194 95 101 707 1570 529 0 55 9 5 12 6 26 119 96 3 0

844 378 42 2640 223 201 1335 3646 1027 2 106 12 9 28 23 62 245 345 4 7

284 149 16 1017 73 95 543 1350 382 0 39 5 4 4 7 25 87 142 1 4

2403 1100 122 7565 625 607 3952 10458 2924 6 311 39 27 71 62 175 703 723 13 14

1 5

2 7

2 5

2 4

1 2

7 18

26 31

35 51

21 18

27 61

24 14

107 144

3

11

4

7

11

33


Written evidence: discussion The diverse written sources of the Vijayanagara period provide abundant evidence of a complex and highly differentiated economy, with an extraordinarily high degree of social and economic differentiation. They also document a rich material culture, including many items that are scarce or non-existent in the archaeological record. As such, they provide a valuable source of evidence on the period that complements the material evidence. Yet, like the later colonial documents, the Vijayanagara period texts are prescriptive. They are valuable in the information they provide on recognized categories of people and idealized landscapes, but they tell us little about how individuals or groups moved among or between those categories, or about the actual material conditions of their lives. We must turn to the material evidence to provide insights into the latter.

A RC H A E O LO G I C A L DATA Along with the many sources of written information on Vijayanagara, we also have abundant, if as yet little studied, material evidence from the period. Many of South India’s major temple centers were constructed or added to by Vijayanagara imperial and regional elites, and temple architecture throughout South India provides the best-documented material evidence of the imperial period (see Michell 1995). The three Vijayanagara capitals – Vijayanagara, Penukonda, and Chandragiri – contain remains of administrative architecture, residences, fortifications, and surface artifacts that can be directly associated with the imperial period. Of these three sites, only Vijayanagara has been the focus of detailed systematic archaeological research. Indeed, little systematic work has been carried out at any major Vijayanagara-period settlement other than its first capital. As a result, most of the archaeological evidence discussed in this work comes from an extremely limited, albeit extremely important, region of the empire and dates to the pre-1565 period. The biases of this restricted geographic focus are significant. While the historic sources are much more widely distributed and help ameliorate the archaeological limitations somewhat, there is a clear need for further archaeological research in broader regions of the empire and at a broader range of sites. Here, I focus on the research that has been conducted at the first Vijayanagara capital and its c. 450-square kilometers hinterland.

139

140


Vijayanagara and its hinterland Much of our knowledge of Vijayanagara and its region has derived from several recent archaeological projects. These include excavation projects carried out by the Archaeological Survey of India and the Karnataka State Department of Archaeology and Museums. Detailed documentation projects of surface remains in the Vijayanagara urban core have been conducted since 1979 under the auspices of the Vijayanagara Research Project directed by John Fritz and George Michell. The Vijayanagara Metropolitan Survey project (1988–1997), directed by Kathleen D. Morrison and Carla M. Sinopoli, has documented approximately 740 archaeological sites in the capital’s c. 450-square kilometers metropolitan region. The city of Vijayanagara is located along the Tungabhadra River, a tributary of the Krishna, in the semiarid uplands of central Karnataka (modern Bellary and Koppal districts). The Sangama founders of the city selected a relatively open valley amidst the rugged landscape of imposing granitic hills. The valley, defined by a bend in the river, provides a valuable pocket of arable land, farmed with the benefit of canal irrigation and a complex system of upland reservoirs, terraces, and wells (see below and Morrison 1995). Given the semiarid climate, with an average of only 50 cm of rainfall per year (and that unreliable, Morrison 1995), access to irrigable land was a key factor in making the region habitable by large populations. Further, the area’s formidable terrain provided security and abundant building materials for the city’s inhabitants (see Brubaker in prep., for a discussion of Vijayanagara’s defensive infrastructure). The Vijayanagara region was sparsely occupied in the centuries prior to the founding of the empire. Although, as discussed in chapter 4, the area had moved in and out of the authority of a range of states, except for the short-lived Kampili state (based approximately ten kilometers north of Vijayanagara urban core), the region had not been a major focus of settlement or of political or economic investment. A number of pre-Vijayanagara temples, mostly attributed to the Chalukya period, attest to some resident population from at least the ninth century AD (and numerous prehistoric remains provide evidence for much earlier human settlement of the region) but, by and large, settlement of the tenth to thirteenth centuries lay outside of the area that would become the center of Vijayanagara-period settlement (Morrison, in prep. a and b). The exceptions to this pattern include a small zone of temple construction on the southern banks of the Tungabhadra River (modern Hampi) and the fortified town of Anegundi to the north of the river (Sugandha n.d.). Several other pre-Vijayanagara temples


and fortresses are located north of the river, as was Kummata durga, the fortress of the Kampili ruler, Kampili-raya (1320–1329 AD; see Patil 1991a, 1991b; Brubaker in prep.; chapter 4, this volume). It is difficult to come up with a precise estimate of the metropolitan region’s population before the mid-fourteenth century. However, it was probably no more than 10,000, and certainly many times smaller than that of the massive urban center that grew with extraordinary rapidity after the empire’s founding in the mid-fourteenth century. Because of the rapid abandonment of the Vijayanagara capital in AD 1565 and the absence of significant reoccupation of the site, preservation of archaeological features and evidence for urban plan is excellent. Archaeologists have conceptually divided the Vijayanagara into several spatial zones, including the broad Metropolitan Region and the urban core and its components. Sinopoli and Morrison (1992) have defined the Vijayanagara Metropolitan Region as encompassing an area of approximately 450-square kilometers. At its center lies the Vijayanagara urban core. This fortified core area encompasses the densest zone of settlement and the city’s major administrative and sacred architecture and has been the focus of intensive documentation of the Vijayanagara Research Project (VRP).

History of archaeological research The first mention of “archaeological” Vijayanagara dates to AD 1567, shortly after the city’s abandonment, when the Italian traveler Cesare Federici wrote: The Citie of Bezeneger is not altogether destroyed, yet the houses stand still, but emptie, and there is dwelling in them nothing, as is reported, but Tygres and other wild beasts. (quoted in Sewell 1900: 208)

The earliest formal archaeological documentation of Vijayanagara dates to c. 1800, when Colin Mackenzie, who later became the first Surveyor General of India, produced a map of the ruins and wrote a description (still unpublished) of site layout and some of the major standing structures (Michell 1985: 196). Several watercolor paintings were also made at this time. Throughout the nineteenth century, other British colonial visitors left brief records of the site. In 1855–1856, Colonel Alexander John Greenlaw came to Vijayanagara with a daguerreotype camera; in the 1980s, the surviving photographs were rediscovered at the home of one of his descendents in Cornwall, England. The collection, now published (Nagaraja Rao 1988), consists of more than sixty 18×16 inch waxed-paper negatives, and provides

141

142


a valuable record of architectural remains at the capital, some of which have since disappeared. Documentation of monuments at Vijayanagara continued in the last few decades of the nineteenth century, as architectural drawings were made of several structures at the site and lithic inscriptions were recorded. In the early 1900s, officers of the Archaeological Survey of India (ASI) undertook a number of clearing and restoration projects at Vijayanagara (summarized in Michell 1985: 200–202); this work continued sporadically into the 1950s and 1960s. Much of the ASI’s work focused on constructing stone and concrete supports for structures under risk of collapse, and sealing roofs and surfaces to prevent further water damage. While this work was valuable in preserving standing monuments, clearance work at many structures resulted in the removal of sediments and archaeological deposits without systematic recovery of artifacts and other materials contained therein. As a result, it is now difficult to interpret the activities that occurred at many of the major structures that remain at the site. The first century and a half of work at Vijayanagara was sporadic and restricted to major monuments at the expense of more mundane archaeological remains or understandings of the broader city plan. Intensive and systematic archaeological research focused on the full range of archaeological materials at the site only began in the late 1970s under the projects mentioned above. As a result of this new work, scholars now have detailed understandings of city plan, chronology, the functions of individual structures or zones, and to a lesser extent, of Vijayanagara material culture, particularly ceramics (Sinopoli 1993b).

The Vijayanagara city core The “city core” of Vijayanagara comprises the zone of the capital’s densest settlement, administrative architecture, and major temple centers. This core region extended over roughly 15-square kilometers in the fifteenth century and nearly twice that area by the early sixteenth century, when the city attained its maximum population. Fritz, Michell, and Nagaraja Rao (1985) have defined three major spatial components of the Vijayanagara urban core. They termed these the “sacred center,” “the royal center,” and the “urban core” (Figure 5.3). I use the term “urban” or “city core” somewhat differently to encompass this entire zone. The area that Fritz and colleagues termed Vijayanagara’s sacred center stretches some three kilometers in a narrow band along the southern bank


T u

ngabhadra River

al

Sacred Center

can

Royal Center

0

km

1

5.3 The Vijayanagara urban core

of the Tungabhadra River. Four large temple complexes were constructed in this area throughout the Vijayanagara period. At the western boundary of the sacred center is the large Virupaksha (Shiva) temple complex; at its east is the ornate late Vijayanagara Vitthala temple complex (dedicated to Krishna, an avatar of Vishnu). A large complex dedicated to Krishna was located to the southeast of Virupaksha, and the Tiruvengalanatha temple (Figure 5.4; often called the Achyutadevaraya temple after the ruler who sponsored its construction) was located near the center of the zone, near the base of the sacred Matanga Hill. Each of these temples consists of a central shrine and associated structures enclosed within a large rectangular walled compound, ranging between 1 and 1.5 hectares in area (Mack 2002). The compounds

143

144


5.4 The Tiruvengalanatha temple

were entered through massive towered gates or gopura – a defining feature of Vijayanagara temple architecture. Extending outwards from the main gate of each walled temple enclosure was a long street, flanked on both sides by columned structures and residential areas. These streets served both as “chariot streets,” along which portable images of the major deities of their respective temples were taken on procession in large wooden chariots, and as bazaar streets, serving the many pilgrims and devotees that visited the temples. Food stalls, guest houses, and stalls selling temple offerings and “souvenirs” would have been present, as well as residential areas for artisans, priests, and temple servants. Direct archaeological evidence for craft production, in the form of production debris or facilities, is not well preserved in these zones. However, Mack (2000) has identified an area of grinding stone production near the Vitthala temple. At the very least, the numerous sculptures and carved columns at the large complexes and numerous smaller structures in the area provide evidence for sizeable numbers of stone quarriers, masons, and sculptors in this zone during periods of temple construction. Written sources also indicate that many different categories of producers were often associated with temple centers (Nagaswamy 1965; chapters 6 and 7, this volume), including potters, weavers, carpenters, stone carvers, as well as bards and musicians.


A large irrigation canal, the Hiruya kaluve (“big canal”) or Turtha (“swift”) canal, separates the zone of temples from main residential zone of the city (Fritz et al.’s urban core). Several inscriptions are associated with the canal, though most are undated. Nonetheless, Morrison (personal communications) has argued that the canal was constructed either in the beginning or middle of the fifteenth century. An undated inscription appears to refer to its actual construction, recording the name of an individual involved in the removal of a section of outcrop.1 This individual, named Karajaunayya, has been interpreted by Davison-Jenkins (1977: 65) as an “engineer.” The approximately 12-square kilometer walled residential center of the city to the south of the canal is an irregularly shaped zone enclosed by massive fortification walls that are pierced by a number of major and minor gateways. Dense areas of settlement were present in lowland valleys between the zone’s outcrop hills, evidenced both by remains of stone masonry architecture and scatters of ceramic sherds. On the hills overlooking these valleys are numerous temples, shrines, watchtowers, walls, and other features (see Fritz, Michell, and Nagaraja Rao 1985: 10–12). Several discrete residential zones are evident in the urban core, of which the best documented is the region contemporary scholars call the Islamic Quarter, containing the remains of a mosque and several tombs. This quarter is believed to have housed some of the Muslim soldiers who served in the Vijayanagara military. An inscription on the mosque dates it to the reign of Devaraya II, the ruler who restructured Vijayanagara’s military through his extensive use of foreign mercenaries (see p. 79). The occupations of the residents of the city’s other residential districts are, unfortunately, not possible to discern. Despite occasional finds of iron slag and carved or flaked stone debris, there is no convincing archaeological evidence that zones of craft production were located within the walls of the urban core, at least for those crafts expected to yield identifiable archaeological remains. However, several of the texts discussed above suggest at least some artisans resided within the city walls. In the southwest quadrant of the walled residential area lay a smaller walled zone, that Fritz et al. have called the Royal Center. This area contains numerous elite residences and administrative structures, and most likely encompasses the palace area described by Paes as “surrounded by a very strong wall like some of the others and encloses a greater space than all the castle of Lisbon” (Sewell 1900: 254). Also, within this zone were broad streets that likely contained the markets that many travelers described as filled with pearls, precious stones, valuable textiles, and other elite commodities.

145

146


The Vijayanagara metropolitan region As discussed in chapter 4, the Vijayanagara empire and its capital grew with extraordinary rapidity in the early decades of Sangama rule. This growth is evident in the city core, where the layout of the royal center was likely completed by the end of the fourteenth century, as well as in the immediate hinterlands of the capital. In Nicolo Conti’s recounting of his entry to the city in the early 1400s, he observed, “the circumference of the city is 60 miles: its walls are carried up to the mountains and enclose the valleys at their foot, so that its extent is thereby increased” (quoted in Major 1857: 6). Similarly, in 1442 Abdul Razzaq described a vast city enclosed within seven “citadels and the same number of walls,” of which the innermost is the palace. He observed: Between the first, second, and third walls are orchards, gardens, and buildings. From the third through the seventh it is very crowded with shops and bazaars. (quoted

in Thackston 1989: 308)

Razzaq estimated that the distance from the northern to the southern edge of the outer fortress was approximately 2 parsangs, or between 22 and 26 kilometers. Using the numerical values provided in those accounts, Conti’s Vijayanagara covered an area of roughly 740-square kilometers, while Razzaq’s city extended over an area of c. 520-square kilometers. As documented in historic accounts and archaeological research, this vast area was a fortified zone of agriculture and settlement. It is this zone (albeit defined somewhat differently) that Morrison and I have termed the “Vijayanagara Metropolitan Region,” and which has been the focus of our archaeological research – the Vijayanagara Metropolitan Survey (VMS) project – a program of systematic survey and documentation. We defined the Vijayanagara metropolitan region as the area contained within the outermost fortifications of the capital. However, in setting the bounds of our survey region (Figure 5.5) we took into account both natural and cultural features. The natural features were topographic – particularly the steep and dramatic ranges of outcropping granitic hills north of the Tungabhadra River that defined the northern boundary of the survey region. The town of Anegundi, which pre-dated and post-dated the empire, is included in the survey region, and is considered as part of the Vijayanagara urban core. The rugged, metamorphic, iron-rich Sandur Hills defined the region’s southern boundary.

5.5 The Vijayanagara metropolitan region (hatched line includes the Daroji Valley, a focus of sixteenth-century expansion)

148


Cultural features used to define the survey region included both Vijayanagara-period constructions and more recent ones. To the west and southwest, the modern Tungabhadra Reservoir and city of Hospet (founded as a Vijayanagara suburb) have largely obscured Vijayanagara-period features. The reservoir, completed in the 1950s, flooded an area of some 380square kilometers and destroyed more than 200 villages. While Vijayanagaraperiod constructions no doubt were present in the now flooded region, sixteenth-century Vijayanagara writings concerning the construction of a suburb near modern Hospet make clear that this was considered the outer boundary of the city. The eastern and southeastern boundaries of the region proved the most difficult to clearly delimit, as here the terrain is more open and traces of Vijayanagara fortifications extend out for a considerable distance. In the southeast, both hill ranges and numerous fortification walls bound the Daroji Valley. The valley is the site of a complex system of interlinked Vijayanagara-period irrigation reservoirs that has been studied by Morrison (in prep. b). A conservative estimate of the metropolitan region that treats the Daroji Valley as a separate topographic unit yields an area of approximately 450-square kilometers; adding in the valley, the Metropolitan Area extends over approximately 650-square kilometers. This last estimate is roughly midway between those derived from the accounts of Razzaq and Conti. Some Vijayanagara constructions, including temples, fortifications, reservoirs, and other irrigation works are found beyond the area we have defined as encompassing the metropolitan region. Nonetheless, site densities, and the pattern of fortifications and other archaeological features, suggest that the survey region we defined was culturally and historically meaningful. Systematic archaeological survey was carried out in the metropolitan region from 1988 through 1997. Because of the high site density and difficulty of the terrain, only a sample of the survey area was covered (see Sinopoli and Morrison in press, in prep., for detailed discussions of survey methodology). Our research focused most intensively on an area of about 160-square kilometers immediately surrounding the Vijayanagara city core, where a 50 percent sample was surveyed and 657 archaeological sites were identified. Beyond this area, more extensive problem-oriented survey was carried out, and an additional 79 sites were documented. Site distributions in the intensive survey area are illustrated in Figure 5.6. Archaeological sites recorded by the VMS fell into a broad number of functional categories. Many had multiple functions; for example, settlements containing wells, gates, or temples. In Table 5.5, I summarize primary site function; more detailed analyses have taken into account the full range of

Vijayanagara: sources of evidence s p ta

a

149

s s r r

d dr a r d

t tr

p t

p s d r

a t ana l

a a s Anego ss s s

p

t

a a a

a

a

t

s

t

rp

d

pp

r

p

d r r rd r r d p s

a

s

a

a

s ss

a

s

s

a t

na

a l

t

l Turtha Cana

c

s a t

s

Kalaghatta

t

s

s

t

s

s

s

a i s a a r aa r r r r

Urban Core

s d s

s

d c s a d

a s ss

s d t t d t

t

t

t

t s

r

t

r

c

s

d

rs

s t d i

a s ss

s

dd s s

s

a

s

a d

a

a

t

a

i

a

a a

Kamalapuram Reservoir

s

a

ss

r

d

c

s s s t

t

m

Raya Canal

i

s

a

a a a a a s a

a ar r r t a a a at rd r a

s

r

s

r

t

t

r r r dCa

s

t

a

s t

t

t

a

s

r

r r r s t s s s s s s sss s r s r ss s s t s a t r i a d d s s s s s ss s s sa r r sd s rr st d s s rr as r a s t s d s s t a a s r s c s a s d a a s a r t s s a a a s a s i a a a a a a a a d a a d a a ad a t ra a t a a a a a aa a a a r a a a a a

r

t

i

a

r

a

r

a

a a

a d a

a s

s

s

s t

a s

m

a

d

a

s

a a

a

t t

s

t s s

r

a

a

a

d

a

t

s

r

t t

t

t

s

t

d

a

a

a

d

a a

d

a

r a

a

d

a

s

a

a

a a d

a

i a a t t

a

a a t t

a

t

t r

a t

a a

a

SITE TYPES s

r

sacred residential

p

pre-Vijayanagara

c civic/ceremonial m mortuary i industrial

agricultural

a t

0

km

5

transport projected transport route d

fortification/defensive Vijayanagara period canals

p p

a

r

s c

a

r a

s

p

s pp s t ap a pt p p a p a t p a d t s d s s d s r sd s s d d ts t a a a s s t s a r s

p

Anegondi

pa p

dr s p p p r p

p

r

s

a

Tu ng a

a

s t

d er a t t Riv dra

d t

bh

p

0 40

s

iC

r pa d r p

s r

nd

d

a

r

s

areas systematically surveyed

Contour interval 20 m Lowest contour 400 masl

5.6 Site distribution in the intensive survey area of the Vijayanagara metropolitan region

150


Table 5.5 Primary site functions documented by the Vijayanagara Metropolitan Survey Primary site function

# of sites documented

1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

161 144 97 77 74 33 9 8 17 69

Agricultural Religious Defense Transport Settlement Prehistoric/Early history Hydraulic (non-agricultural) Industrial/craft production Other Unknown

activities and archaeological features documented at each site (see Sinopoli and Morrison in press). As is evident from Table 5.5, very few sites were classified as primarily being craft production sites. However, while craft production evidence was rare, it was not quite as invisible as the primary site function numbers make it appear. Instead, evidence for productive activities such as stone working and metal smelting and forging often occurred in sites with other primary functions, such as at settlement sites or in association with temples or agricultural features. In total, various kinds of productive activities were documented at nearly 100 sites in the metropolitan region. Evidence of quarrying activities was also noted throughout the metropolitan region; isolated quarrying marks or individual worked boulders were typically not assigned site designations, although their occurrence was noted. I present detailed discussions of the archaeological evidence for craft production in the survey area in the discussion of individual crafts in chapter 6. I focus here on general patterns of land use and settlement in the metropolitan region, and their changes over time. Although we are not able to date many of the sites and structures in the metropolitan region as precisely as we would like, both archaeological and inscriptional evidence confirm that the chronological sequence in the Vijayanagara metropolitan region closely paralleled the political fortunes of the empire as a whole, with rapid growth in the late fourteenth and early fifteenth century under the early Sangamas. During the late Sangama period, there was a marked decline in the rate and scale of construction. Population was at best stable, and may well have declined in the late fifteenth century.


151

urban core

Kamalapuram

Varadadevi amana-pattana

Kamalapuram Reservoir

shrine, temple residential zone

fortification wall

0

km

transport route

agricultural

5.7 Varadadevi-amana-pattana: distribution of archaeological remains

Under the Tuluvas of the early sixteenth century, the pace and scale of construction increased dramatically, and the capital reached its maximal size and population. New settlements were created, at Hospet and to the southeast of the urban core, at Varadadevi-amana-pattana (Figure 5.7), and new fortifications and irrigation works were constructed. From the city’s founding, Vijayanagara’s rulers and elites devoted considerable resources to creating agricultural zones near their capital. As in earlier Deccani states, the construction of agricultural facilities was a major focus throughout the Vijayanagara period, with 157 sites classed as agricultural sites. Canals were constructed in low-lying areas near the river, and reservoirs, wells, terraces, and erosion control features were built throughout the upland zones to facilitate rainfall agriculture. The largest of these constructions were as much as three kilometers long, and required enormous labor forces, ranging from manual laborers to move earth and stones to more skilled laborers, such as architects, engineers, masons, and stone quarriers. As discussed in chapter 4 (pp. 107–108), we know little about how labor was recruited for these and other large-scale construction projects. Among the other large-scale works documented by the VMS was the construction of the city’s defensive infrastructure. Although the widespread belief that Vijayanagara was enclosed within seven concentric rings of fortifications is incorrect, reflecting literary convention rather than actual layout,

1

152


the capital was heavily fortified (Brubaker 2000). Defensive sites include fortification walls, hilltop forts, outposts, isolated bastions, and horse stones. The impressive dry-stone masonry walls of the original Vijayanagara urban core were probably largely completed by the early fifteenth century (Brubaker in prep.). By the early sixteenth century, this core area of dense settlement had expanded through the addition of an outer fortification wall that encompassed the pre-existing walled town of Kamalapuram, Varadadevi-amana-pattana, and the area between them. Vijayanagara’s defensive infrastructure enclosed far more than the city core; indeed, walls and forts bordered the entire metropolitan region (Figure 5.8). Along the northern outskirts of the metropolitan region lay a string of hilltop forts, each with excellent visibility of the others and of the urban core and low-lying regions south of the river. Similar forts are found in strategic locations throughout the region. Elsewhere, fortification walls spanned strategic passes and potential access routes. The builders of these walls were clearly concerned both with their strategic effectiveness and with the visual impact that they would have on those who passed through them. In many areas of the metropolitan region, the walls were constructed of locally available stones, with dry-stone masonry only on the exterior face. In some more peripheral regions, the walls were less formally constructed, built of locally available unmodified cobbles. However, along major roads, walls and gates were usually faced with cut or carefully fitted stones and were often quite massive and ornate, designed to impress and intimidate as well as defend. As with agricultural facilities, the skill and labor required for such constructions was considerable, divided among various categories of specialists and numerous manual laborers. As discussed in chapter 6, the variations in wall form suggest that construction was divided among multiple teams of laborers. Sacred sites comprise the second most numerous category of site types recorded by the VMS, and include large and small temples, shrines, and religious sculptures. These sites embody considerable variation in scale and workmanship – from large elaborate temple complexes (Figure 5.9), sponsored by kings and other elites, to small single room shrines and isolated sculptures of varying quality. Again, a wide range of individuals and production groups were involved in the construction of sacred sites. Other common site types in the metropolitan region include roads, gates and other sites related to transport and movement, and settlements – including large towns and smaller villages and agricultural settlements (see Sinopoli 1997). Wells were often located along major roads, to serve pilgrims and travelers. Thirty-three prehistoric and early historic sites were also documented in

Nagalapur (Hospet)

a dr

River

0

km

5

URBAN CORE

Sandur Hills

Kamalapuram

ANEGUNDI

Contour Interval = 40 Meters

5.8 Vijayanagara’s defensive infrastructure

Fortification wall

Hill fort

Fortified town

Tungabhadra Reservoir

a

Tunga bh

Daroji Reservoir

Vijayanagra

154


5.9 Vaishnava temple complex (VMS-142), Varadadevi-amana-pattana

the metropolitan region (Morrison 1998; Sinopoli and Morrison in press). Many other sites of unknown function (e.g., isolated wall fragments) were also documented. The Vijayanagara Metropolitan Survey provides evidence for a complex, multi-layered, urban landscape. Its inhabitants included soldiers, farmers, nomadic shepherds, administrators and nobles, priests and artisans. The metropolitan region was a dense zone of agriculture, of wet and dry crops,


and it was a heavily protected fortress, defended by thousands of soldiers. The metropolitan region was also inhabited by many of the artisans who built the city and supplied its daily goods and luxury products. I turn to consideration of these artisans in chapters 6–8, incorporating the archaeological evidence from the metropolitan region as well as the diverse written sources of evidence summarized in this chapter. Note 1 The text, carved on a boulder near a small carving of Nandi, is translated as “The person, Karakunayya, donated this. This Nandi rock had to break” (DavisonJenkins 1997: 65).

155

6

Craft products and craft producers

Vijayanagara artisans manufactured an impressive array of craft products, which differed in productive technology, labor investment, and contexts of use. In this chapter, I present information on many of these craft products, and on what we know about the materials, tools, techniques, and geography of their manufacture. I also discuss evidence, derived from archaeological and textual sources, on the social composition, size, and division of labor within craft-producing units and communities. As this discussion will show, there is considerable variation in all of the above dimensions. Within this variation, two general patterns are evident. The first is the high degree of specialization that existed in the vast majority of productive tasks. There is also evidence that this specialization increased, in at least certain crafts, throughout the Vijayanagara period. This trend is most clearly manifest in textile production, which involved technically specialized weavers, dyers, washers, starchers, cotton carders, tailors, and merchants. Such artisans were engaged in diverse and complex interactions that spanned multiple communities and broad geographic regions. The concept of a village artisan working in timeless isolation manifest in the writings of Mills, Marx, and other theorists of the Asian state discussed in chapter 3, does not hold up to scrutiny. Nor, as will become clear, do concomitant claims for despotic state authority or the dependence of artisans on royal households. The second pattern evident in the Vijayanagara data concerns the prominence of households as the primary social units of craft production. As discussed in chapter 4, there is evidence that it was possible to recruit work groups of hundreds or even thousands of individuals for major construction tasks such as building reservoirs, fortification walls, and other monumental constructions. Nonetheless, many craft-production activities occurred in household contexts and involved production units of ten or fewer individuals. Many of the members of these households were craft specialists by any definition of the term. They gained the vast majority of family income from craft production for external consumers, and multiple household members engaged in craft production activities for significant portions of the year. They procured raw materials in a variety of ways, including direct gathering or production or as finished products from other specialists or markets; they


produced goods for markets, merchants, individual or institutional patrons, or to fulfill obligatory economic relations. Yet, even where inter-relations among producers were multiple and complex, it remains the case that most artisans worked primarily in household contexts and often with relatively simple technologies. An important implication of this pattern is that, with the expanding demands for many craft products that occurred during the fourteenth through seventeenth centuries in South India, we more often see a multiplication of the numbers of craft producers and household workshops. For most crafts, there is no evidence for a shift to larger, more technologically efficient, or more hierarchically structured production units. The combination of hyper-specialization in an extremely complex polity with relatively small units of production appears at first glance discordant with the typologies of craft specialization presented in chapter 2 and with conventional social evolutionary thought on organizational change in complex societies. Such models take as their premise a “Fordist” approach – that output is increased through making production more efficient via investment in and expansion of production facilities and increased task differentiation. Labor is seen as a limited good that must be managed efficiently and effectively in order for output and quality of craft goods to be increased. The data from Vijayanagara suggest an alternative scenario, in a context where labor was plentiful, and alternate routes to increasing production were possible. Such a scenario may have relevance to many precapitalist societies (and its lack of fit with early capitalist eighteenth- and nineteenth-century European conceptions of labor organization may also account for some of the interpretation of South Asian stagnancy discussed in chapter 3). In South India, production was increased not by intensification through dramatic technological changes or organizational transformations, but by increasing the numbers of producers. The Vijayanagara ruler Krishnadevaraya may have alluded to such an approach to labor organization in the literary work the Amuktamalyada, a Telugu epic poem written in the king’s voice and believed to have been composed by the king himself (see p. 132): It is desirable to employ several officers to do the work which may be done slowly by a single person; for the work is quickly performed by these and all their dependents. Decrease in the number of officers gives rise to grievance. Increase, on the contrary, causes contentment. (quoted in Nilakanta Sastri and Venkataramanayya 1939:

155)

This discussion should not be taken to undervalue evidence for increasing task differentiation or diversification during the Vijayanagara period. We see this in the appearance of new named occupations, such as cotton carders

157

158


for textile production, where we also see the emergence of a small number of multi-loom production units under the control of “master-weavers” (see pp. 185–187), a rare but significant development in the fifteenth and sixteenth centuries. But by and large, the broader pattern is one of increasing numbers of specialist producers being pulled into regional and political economies through a variety of mechanisms. These include migration and settlement redistribution, processes of cultural incorporation (i.e., what Srinivas 1989 has termed “Sanskritization”), and inducements and privileges, as well as coercion. These economies in turn were, as discussed in chapter 4, characterized by increasing monetization, which had the effect of drawing producers deeper into regional political economies to generate cash revenues for taxation, temple donations, and, possibly, other expenses. I will return to a discussion of these general patterns in chapters 7 and 8. In the remainder of this chapter, I focus on a detailed discussion of the technology and organization of a range of craft products and craft producers in fourteenth- through seventeenth-century South India.

C AT E G O R I E S O F C R A F T S Drawing distinctions between rural and urban or elite and non-elite crafts for the Vijayanagara period is problematic, and such binary terms gloss over a great deal of variability. It is not, by and large, possible to make such distinctions in a straightforward way based on category of material, as many goods produced by specialists took a variety of forms and served a broad range of consumers. Thus, textile production included both the manufacture of finely worked elaborate silks and cotton cloths for affluent consumers and commerce, as well as the production of coarse cotton textiles for the garments of commoners and laborers; iron smiths manufactured both weapons of war and plough shares. It is therefore not possible to speak of all weavers or all smiths as sharing a single identity or a unitary social or economic status. Artisans who worked in villages and supplied the needs of local agriculturalists and other low-status consumers lived very different lives than those who dwelled in urban contexts or temple towns and/or produced for political and religious elites or international commerce, even if they belonged to the same caste or hereditary community. There is thus an important (if difficult to qualify) distinction to be made between goods with high social, cultural, economic, and ideological value and those without, and the experiences and social and economic opportunities available to their producers. These valued goods encompassed a range


of materials. They included the following, among others: swords, weapons, and fittings for military horses and elephants; ornaments and containers of precious stones and metals; elaborate silk and cotton textiles and garments; bronze, stone, and ivory sculptures and figurines of gods and humans; paintings; and even epic poems, produced to honor kings, gods, and other patrons. The producers of such goods stood in distinctive and diverse relations to the consumers of their products and I will explore specific categories of producers and products in some detail below. Here, I restrict myself to a few general comments. Some of the “elite” goods mentioned above were produced for and distributed through markets or other forms of exchange; others were produced for specific patrons. However, even in the latter instance, we cannot assume that the patron–producer relationship necessarily took the form of dominating patron and dependent producer (e.g., Costin’s (1991) individual retainer or nucleated corvée production). Producers, acting individually or as members of a community, often played important and indeed dominant roles in setting the terms of their relations with patrons and in deciding how and whether to participate in them (Shulman 1992). Primary among the factors that afforded producers this ability was the large numbers of potential patrons or institutions with which they could affiliate, including patrons both within and beyond the territories claimed by the Vijayanagara polity. Within the empire, potential patrons included members of the imperial court and elites at the imperial capital, and institutions, particularly the many temple centers, large and small, found throughout southern India. Equally important were the various regional elites, including local rulers, nayakas, military leaders, sectarian leaders, affluent merchants, and others, who benefited from the economic expansion and dynamic political environment that characterized South India of the fourteenth through seventeenth centuries. The intense social competition among diverse kinds and categories of elites and institutions during the Vijayanagara period fueled demands for goods that could represent and symbolize status and prestige, both secular and sacred. Producers of especially valued products were able to benefit greatly from this dynamic political environment and were able, as communities or individuals, to relocate to more advantageous regions or affiliate with more rewarding patrons, as circumstances warranted. Further, demands for elite craft products and skilled artisans did not stop at the boundaries (however fluid and vaguely defined) of the Vijayanagara empire. Recent writings by Eaton (2000), Talbot (1995, 2000), and Wagoner (1996, 2000; see also Sinopoli 2000) have refocused our attention to the

159

160


broader political universe in which Vijayanagara participated, in particular the complicated relations of Vijayanagara with its neighbors to the north, the Bahmani sultanate and its successor states. Architectural forms, clothing styles, and literary works all attest to the complex and ongoing nature of interactions among these polities. These interactions entailed both conflict and diplomacy, and the period witnessed the emergence of at least a partially shared idiom of conceptions of royal power, authority, and representation that spanned political boundaries. Goetz (1940) has suggested that following the 1565 AD abandonment of the empire’s first capital, some Vijayanagara artisans may have moved to the Deccani courts or even further north. Such inter-polity movement of artisans must have also pre-dated the abandonment. Certainly, the architectural connections between Vijayanagara royal administrative architecture (Michell 1992a, 1992b, 1995) and constructions at the contemporary Deccani sultanate capitals of Golconda, Ahmadnagar, and Bijapur suggest that architects and builders may have traveled, along with ideas about appropriate royal architectural forms. It is likely that other kinds of skilled artists and artisans also crossed political boundaries to establish relations with a range of potential patrons. The complicated and fluid political dynamic that characterized South India in the fourteenth through seventeenth centuries may well explain why we seldom see the conventional forms of attached specialization that archaeologists have come to expect in complex societies. In a political climate where potential patrons are many and the products of specialists are highly valued for both their economic value and their role in sociopolitical competition, then the roles and status of producers may be considerably more fluid than in less complexly structured polities. Categories of crafts and categories of consumers and the range of relations between them must therefore all be taken into account in interpretations of, or model building about, the nature and organization of specialized production. In the remainder of this chapter, I turn to specific specialized products to explore some of these issues further. I note at the outset that while I will consider a wide array of crafts in this chapter, the discussion is by no means comprehensive. Instead, I restrict my discussion to those spheres of production for which we have either a significant textual or material record.

Poets and bards It is admittedly rare for an archaeological work to consider nonmaterial “crafts” in a discussion of craft production. Such works are rarely preserved,


even from literate societies, and they fall well outside of the material expertise of archaeology. But in cases where we have such sources, nonmaterial crafts are valuable to explore. This is especially so given that in many societies song, poetry, and laudatory texts are valued “prestige goods” that play a crucial role in ideological communication and processes of political legitimation. They are also highly “transportable” and can be communicated across broad social boundaries. Literary works certainly fulfilled these roles in the Vijayanagara context, where poems and songs were of great importance to courtly life as well as religious ritual. Poetry and song, including satirical works, no doubt also formed an important part of the oral tradition of non-elites, and many anonymous and ascribed poems have been remembered, transferred, and transformed through oral recitation over the centuries (Narayana Rao and Shulman 1998: 3–4). The rich literary evidence from fourteenth- through seventeenth-century South India provides us the opportunity to examine literary works as a valued craft with broad ideological, sociopolitical, and economic impact. As discussed in chapter 5, many major poets and their works are known from the period. Often their poems are self-referential, so that we learn both about poetic subjects and patrons (always described in a poem’s opening passage, often in highly formulaic ways) and also about their authors. Further, allusions to goods or categories of people that appear in poems often provide valuable information on the cultural values of other craft products or their producers. Unlike many of the other crafts discussed in this chapter, the production of literary works often involved individual artists whose names and biographies are known to us. The most honored of these artists – renowned poets attached to imperial courts or great temples – obtained positions of considerable status, and many became enormously wealthy and influential. Also, unlike other crafts, poets could come from a wide range of hereditary communities, though the “ideal” poet was a high-status male Brahman who was both poet and scholar (Narayana Rao 1992: 144). Poets from the Kshatriya and Shudra varnas are, however, also known, and non-specialists sometimes engaged in the craft of poetry writing. Military leaders, kings (e.g., Devaraya II, Krishnadevaraya), queens, and royal princesses are attested to have composed poems and songs (see chapter 5, pp. 130–135). Common poetic subjects of the period included the rewriting of ancient Sanskrit tales and sacred texts, stories of romantic love, and recountings of historic events and heroic figures. Poets also sometimes composed texts for copper plate and lithic inscriptions recording royal donations; the texts of at least five copper plate

161

162


inscriptions that record donations made by Krishnadevaraya are listed as having been composed by an individual named Sabhapati, described as the author of “pleasant and graceful words” (EC VIII 1984: 585–587). Sabhapati also composed the texts of copper plate inscriptions that record donations made by Achyutadevaraya and Sadashivaraya. A grant from AD 1636, now stored in the Kamalapuram museum near Vijayanagara, records a royal gift made by the king to a Brahman. It notes that the text of the grant was composed by Rama-kavi, who describes himself as the son of Kamakoti and the grandson of Sabhapati, suggesting that in this family, poetry was an hereditary occupation (ARIE 1972–3, A5, summarized in Patil and Patil 1995: 128). The poet Nrihari is recorded as the composer of “faultless verses” on an AD 1639 copper plate recording a donation to Brahmans made by the Aravidu ruler Venkatapatirya II (EC III 1984: 711–714). Other composers of copper plate and lithic inscriptions are sometimes named in the inscriptions. Along with people explicitly designated as poets, composers of inscriptions included temple accountants (senabova) and others whose primary occupational identifications are not mentioned. South Indian poetry of the Vijayanagara period was composed in several languages, including Sanskrit, Tamil, Kannada, and Telugu. Sanskrit was particularly important in temple and courtly contexts and Tamil poetry had a long and esteemed literary history dating back to the Early Historic Çankam period of the early centuries AD (e.g., Shulman 1992). Poetry in Telugu underwent a great expansion during the early sixteenth century. Telugu poets were heavily patronized in the court of Krishnadevaraya, who was renowned as a great patron of the literary arts and was himself reputed to be a poet of no little worth. As noted above, he is the acknowledged author of the Amuktamalyada and other works in both Telugu and Sanskrit (see Shulman 1995). His selection of Telugu was interesting, as he is believed to have been a native Kannada speaker (Raghavacharlu 1936). A passage from the Amuktamalyada recounts a visitation that the king received from the god “Andhravishnu.” The god commanded Krishnadevaraya to compose a tribute to him in Telugu saying: “You have heard many languages from the kings who came to serve you, and you know: Telugu is best among the country’s languages” (quoted in Narayana Rao 1992: 145). As a Kannada speaker and a poet who had already demonstrated his proficiency in the sacred language of Sanskrit, Krishnadevaraya’s use of Telugu is an acknowledgment of his relation to his Telugu subjects and may have lent support to his status as a universal ruler. The universal character of Krishnadevaraya’s court and its support of poetry were proclaimed in many sixteenth-century literary works, and


remained a powerful trope in post-Vijayanagara literature and popular thought from the Nayaka through contemporary periods (e.g., see Wagoner 1993; Narayana Rao 1992; Shulman 1995). A poem by Allasani Peddana, one of the great Telugu poets patronized by Krishnadevaraya, described the king’s court: Women with faces like the autumn moon, their eyes shining like dark lotuses, were waving chowries and scholars were discussing the details of the sutras of Kanda, Panini, and Badarayana while guards stood nearby, fearless men with terrifying swords of blinding brightness and tributary kings surrounded him with the soft light radiating from their brilliant gems as the glittering rays from the jewels on his war anklet that marked him as ‘master of the Three Kings’ were dancing all around the borders of his white silk garment which then seemed to be of many quivering colours. While he sat in splendour in his assembly hall called “Conquest of the Universe,” discoursing with his scholars, his heart was moved by the sweetness of poetry and, from his throne, he spoke to me with kindness. (excerpted in Narayana Rao 1992: 150)

While kings may have been the most prominent patrons of poets, they were by no means their only patrons. Indeed the multiplicity of potential elite patrons contributed to dramatic changes in the dynamics between artists and their sponsors, and to an increased diversity of poetic forms. David Shulman (1992, 1995) and V. Narayana Rao (1992) have explored the complex and mutually legitimating relations between poets and their many possible patrons. Potential “secular” patrons included kings, military leaders, ministers, various regional subordinates, and even merchants and village leaders (Narayana Rao 1992: 143). This contributed to a marked hierarchy of poets – linked to the status of their patrons and the reputed quality of their words – but also contributed to the poets’ mobility and their potential to manipulate patrons in order to secure greater rewards for their service. Other poets were attached to temples or deities, and acknowledged the gods as their sovereigns. Still others were associated with devotional bhakti cults. In many respects, the structural positions of temple poets paralleled

163

164


those of courtly poets. However, while courtly poets were awarded great economic and social privileges (such as the right to be carried on a palanquin, have an umbrella held over their heads, and wear pearl necklaces and garments embroidered with gold; Narayana Rao 1992: 156), temple poets were idealized as renouncing such luxuries and living a life of pure and impoverished devotion (though this did not stop some from becoming wealthy, see pp. 279–280). Temple poets frequently depicted themselves as resisting seductive entreaties or coercive pressure from kings to shift their allegiance away from their divine patrons, and were at times highly critical of royal authority. This is particularly evident in the writings of Dhrujati, a sixteenthcentury poet affiliated with the large Kalahasti temple center in Andhra Pradesh (a center that received considerable patronage from Vijayanagara rulers Krishnadevaraya and Achyutadevaraya). Dhrujati wrote: Those kings drunk with power, serving them is like being in hell. The things they give you – women with eyes like lotuses, palanquins, horses, jewels – all breed pain. I’ve had enough of wanting them. O God of Kalahasti, through your grace change me so that I awaken to the wealth that is Illumination. (translation in Heifetz and Narayana Rao 1987: 20)

And: If those king who are thorns for the world were even to offer me ten thousand pleasures, my mind would disdain them, but show me a king who is just and truthful and compassionate, and I will regard him as I regard you, O God of Kalahasti, and feel happy at the end of every day. (translation in Heifetz and Narayana Rao 1987: 103)

By the fifteenth century we also see evidence for wandering poets, who appear in both Tamil and Telugu literature (Shulman 1992: 107). These


authors formed short-term relations with a variety of patrons, and moved widely across the South Indian landscape – serving to convey information between various courts and patrons, and actively participating in the broader ideological and political changes of the period. They added a new dimension to existing patterns of poet–patron relations (Narayana Rao 1992; Shulman 1992, 1995). While poets required patrons, it is also the case that rulers and would-be rulers needed poets to legitimize their positions. Narayana Rao (1992: 146) expressed this relation as follows: “Great poets were those who lived in the courts of great kings. Great kings were those who had great poets in their courts.” The patronage of poets provided sponsors with “symbols of status” (Narayana Rao 1992: 143) that were part of the complex webs of social manipulation and competition that characterized the period. As a result, poets had considerable power in these competitive dynamics. This is not lost in the content of the poetry, and during the Vijayanagara and subsequent Nayaka periods (when there were even more claimants to political authority), new poetic forms and themes appeared. In this new genre, kings were often portrayed as inferior to poets, who use their words and their wiles to outwit rulers and enhance their own status (Narayana Rao 1992; Shulman 1992). That some poets became extraordinarily wealthy is well attested from both poetic descriptions and inscriptional sources. Several inscriptions from the Tirupati temple refer to donations made by poets. For example, in AD 1540 the temple poet Sri Chinna Tirumalaiyangaru made a donation for the purpose of reconstructing a temple that had fallen into ruins in the nearby imperial city of Chandragiri and to support daily offerings in the temple (TTDES IV, GT429, p. 266). In 1541, the poet Tallapakkam Periya Tirumalaiyangara made a cash donation to support ritual offerings that were to be made within the Tirumalai temple complex near the room where the copper plates of poems were stored. This offering consisted of 450 panam (coins of a copper and gold alloy, see chapter 4) and a grant of the income from two villages, reported to generate 300 rakhai-pon (gold coins) per year (TTDES IV, TT682, p. 287). In return the poet would receive rice and prasadam (a share of the consecrated temple offerings) each day. The famous fifteenth-century temple poet of Tirumala temple, Tallapaka Annamacarya (1424–1503), who reportedly composed more than 32,000 works dedicated to the Lord Venkatesvara, also made large donations to the temple (Narayana Rao 1992: 155). In the case of poetry and literary works, we see evidence for a mode of “craft production” characterized by patron–client relations, a variant of

165

166


attached specialization. In this variant, specialists had considerable sway in their relations with their patrons, whom they could elevate or humiliate through their words. However, while court poets benefited greatly from this relation in terms of their wealth, status, and renown, they could also be brought down. This occurred most dramatically when rulers were weakened or overthrown, and several of Krishnadevaraya’s most esteemed court poets are reputed to have ended their lives in poverty following their patron’s death (Narayana Rao 1992: 155). Less dramatically, a court poet’s status could decline as patronage shifted to other poets within the court. As a result, competition among such artists was extreme and unlike other artisans discussed below, there was little sense of a collective good or collective action among elite poets. Beyond the courtly sphere, there was a more subversive dimension to the many of the period’s literary products. This is exemplified in the writings of devotional and temple poets such as Dhrujati, who expressed a profound distrust of all political authority and royal claims; a distrust that probably existed as well among wandering poets and village bards whose words were not recorded. Temple songs and wandering bards provided contexts in which both of these conflicting messages of South Indian poetry were widely disseminated.

Performers: music and dance Another dimension of craft production seldom discussed in archaeology concerns the performing arts of music and dance. The material correlates of such crafts include instruments, ornaments, and theatrical spaces. For Vijayanagara, our major sources are literary – texts and inscriptions – and representational, particularly sculptures and friezes depicting performance (Verghese 2000). It is clear from these diverse sources that music and dance were extremely important components of courtly and temple activities throughout the Vijayanagara period. They no doubt figured as well in the lives of non-elites, though the sources for exploring this are scarce. Here, I first briefly discuss courtly performance and then turn to a consideration of temple performers, including the devadasi (“servants of the god”) – unmarried temple women, some of whom accumulated considerable wealth and status during the Vijayanagara period. Both male and female performers appear frequently in literary works and foreigners’ descriptions of the Vijayanagara court. Abdul Razzaq (in Thackston 1989: 314) described a performance he observed in AD 1443


during Vijayanagara’s annual Mahanavami festival (when tributary lords came to the court to swear their fealty and pay tribute to the emperor): A hundred female singers, a crowd filled with moons and Jupiters When they removed the veil from the sun, with one wink they destroyed the world With countenances shining like the sun, they scorched people’s souls with their fire. When the veil was dropped from their faces, the moon went veiled into a pit. When they danced, their statures were musk locks sweeping the ground When they struck the ground with their dancing feet, they kicked Venus from the competition. The audience were all astonished by their motion and movement. Orr (2000: 178), summarizing Pamela Price, has observed that by the Vijayanagara period, women of the court, including performers, had become an essential component of courtly ritual and presentation, displaying the qualities of “auspiciousness, material well being and fertility” that were essential to representations of kingship. As with poets, female and male performers were necessary to the expression and maintenance of royal status at both imperial and regional courts. We do not, however, have the evidence to determine whether or not performers had the same degree of social mobility that successful poets did. It is certainly possible that some individuals whose skills were highly valued were able to define their relations with patrons. But the majority of performers were likely from lower-status hereditary communities or, in the case of court or temple women, may have had very little control over their labor and its location. An inscription from the Venkateshvara temple at Tirumalai, for example, refers to a temple woman named Hanumasani, the daughter of Uddida Timmayyan, who is described as having been sent to the temple by the Vijayanagara ruler Achyutadevaraya. Hanumasani was nonetheless in the position to make a cash donation of 820 nar-panam (coins in alloy of gold and copper, see chapter 4) to the large Venkateshvara at Tirumalai and have it recorded in an inscription (TTDES IV, TT422, p. 263), indicating that while she could not exercise choice over where she lived, she did have control over personal wealth. Sculptures of female performers are common at Vijayanagara, appearing on sacred and civic-ceremonial architecture (Figure 6.1; Verghese 2000). Women, dressed in a variety of styles, are depicted dancing and playing

167

168


6.1 Female performers on the basement of mandapa in the Vitthala temple complex, Vijayanagara Sacred Center

drums and stringed instruments. Male performers also are sculpted. Several such sculptures occur on a large elevated platform at Vijayanagara that is called the Mahanavami or royal throne platform. These representations depict distinctive bearded men playing drums and dancing (Figure 6.2). These individuals appear to be foreigners, and Verghese (2000: 285) has suggested that they may be Central Asians. Musical instruments described in texts and illustrated in sculptural representations of performers include drums, flutes, trumpets, and various stringed instruments. Music also played an important role in temple worship. Musicians, particularly drummers, often appear in lists of temple functionaries. Images of dancers also appear on temple carvings. Vocal music was also important to both temples and palaces, and the most famous composer of Vijayanagara was Purandaradasa, who resided at the capital during the reign of Krishnadevaraya in the sixteenth century and is reputed to be the founder of contemporary Carnatic classical music. He and other great composers were most similar to the elite poet bards discussed earlier, in their status and social position.


6.2 Bearded male performers on royal throne platform, Vijayanagara urban core

We know relatively little about how performers were recruited. Most likely came from hereditary communities of performers. Some may have been from rural or tribal communities. Depictions of the “stick dance,” (Figure 6.3) believed to have derived from local folk traditions, occur on the royal throne platform and on the basements of several palaces. Others, particularly the female performers in temples and courts, may have also been recruited through non-hereditary routes. Temple women were the subject of frequent commentary by eighteenth and nineteenth-century colonial administrators, who viewed the devadasis (servants of the lord) primarily as exploited dancers and prostitutes (see discussion in Orr 2000). While these women were believed by these authors to have had a more noble status in earlier times (i.e., as holy temple virgins), colonial writers argued that their position had become debased over the centuries (as indeed had Indian society overall, see discussion of the degenerative view of Indian history in chapter 3; also Orr 2000: 3–4). As Leslie Orr (2000) has recently discussed, this generic and highly problematic view of temple women has been extended to other periods and places in South Indian history, ignoring historical particularities and temporal changes. In marked contrast, Orr has conducted a detailed analysis of inscriptional sources concerning temple women from the Tamil region during the pre-Vijayanagara Chola period. These unmarried women appear to

169

170


6.3 Female performers sculpted on royal throne platform, Vijayanagara urban core: the stick dance is depicted in the upper panel

have had various associations with the temples. Orr concludes that during the late Chola period, temple women (usually referred to as “devotees of the lord”), most often appear in the inscriptional record as donors to local temples. Typically, these are small local temple centers, located on the margins of Chola imperial power, where state power was weak or contested. This suggests to Orr that it was within these more fluid political contexts that some Chola period women were able to attain significant positions within temples. Their status depended on their actions as donors, rather than on being recipients or beneficiaries of temple support. During this period, temple women do not appear to have been associated with particular ritual or administrative functions; they were not defined specifically as dancers or in terms of any other specialized role. After AD 1300, the roles and positions of temple women in the Tamil region seem to have changed in a number of respects. This includes a shift toward a definition of temple women as a hereditary and professional specialization, though the precise tasks these women engaged still remain remained largely unspecified (Orr 2000: 175). Temple women continue to be unmarried, and their descent is often reckoned matrilineally. However, in the post-1300 period, Orr found evidence for more formal relations and alliances between temple women as a category of people and other artisan


communities, such as the Kaikkolar weavers. This may have contributed to (or resulted from?) a weakening of the women’s individual autonomy. Her analysis suggests that by the Vijayanagara period, temple women increasingly appeared as recipients of temple support, rather than as donors, suggesting significant changes in their relations with institutions. Finally, Orr notes that there was increasing congruence between temple women and women of the court, and perhaps, as noted above, a changing symbolic value attributed to femaleness as an essential accompaniment of royal and temple performance. In this work, I consider temple women in the category of artisans because of their role as dancers, which appears to have become increasingly important over time (Orr 2000: 175–180). This is not to deny the other roles and positions of these women. Temple women appear in fewer than a dozen inscriptions in my database of inscriptional references to craft production. In approximately half of these, the women are noted as recipients of the proceeds of temple donations. For example, an inscription from a temple in Kagaiyur, Tamil Nadu dating to AD 1524, records a donation by an agent of Tirumalai Nayaka and local elites in which revenues from donations of land and loom taxes were allocated to the temple for the support of three weaver chiefs (Kaikkolar mudali) and fourteen temple dancers who performed services in the temple (Karashima 1992: 48). In Gundlupete, Mysore District in AD 1382, the smith Vira-pancha along with other smiths and braziers of named villages made a cash donation to be used for the support of the temple woman Ketavve (EC III 1974, p. 568–569). In the remaining inscriptions (for example, the donation by Hanumasani discussed above), temple women appear as donors, suggesting that, while there may have been structural changes in the status of temple women, individual women retained access to wealth and resources.

Weavers and textile manufacture Perhaps more than any other material craft, investment in textile production – in labor and resources – can span a nearly infinite range. Undecorated cloths of coarse thick yarns can be formed relatively quickly using simple technologies with little per piece labor investment. In contrast, cloths of fine yarns made of valued raw materials and in elaborate patterns may take hundreds or even thousands of hours to complete. Cloth in its infinite variety can thus become a potent medium for symbolic expressions of status and wealth (see Weiner and Schneider 1989; Barnes and Eicher 1992;

171

172


Hendrickson 1996). Add to this the visual prominence of textiles and their close associations with the human body, and it is not surprising that cloth has played a key role in the display and communication of social information in South Asia (Joshi 1992; Cohn 1989; Wagoner 1996) and throughout the world. During the Vijayanagara period, the intensity of textile production increased dramatically to serve the demands of expanding elites and changing styles of royal dress, as well as of expanding internal and maritime commerce (Subrahmanyam 1990a). South Indian textiles were highly valued in Europe and throughout Asia. Centers of textile production were often located at major ports and along transport routes (Ramaswamy 1985c). The importance of cloth production during the Vijayanagara period contributed to the prominence of weavers in a variety of contexts, particularly in temples. State recognition of weavers also occurred, as the social and economic status of weaving communities improved considerably throughout the period. I will explore the relations between weavers and institutions in more detail in chapter 7; here, I present information on the composition and geography of textile-producing communities, productive technologies, and textile use. I conclude with a brief discussion of a community of weavers that rose to considerable prominence during the Vijayanagara period – the Kaikkolar weavers of the Coromandel region of southeast India. Because of the non-durable nature of textiles and of most of the tools associated with their manufacture, direct material evidence of Vijayanagaraperiod cloth production is virtually non-existent. Sculptural representations of dress styles and rare Vijayanagara-period paintings do provide some material information on textiles and temporal changes in clothing styles. In contrast to the scarcity of material evidence, textual information on weavers and their products is comparatively abundant. Textile producers are mentioned in numerous temple inscriptions, where they appear as donors and temple functionaries, as taxpayers and recipients of tax abatements, as holders of a range of social and ritual privileges, and as participants in other legal disputes and social protests. Poems and literary sources of the period frequently describe elaborate garments and named categories of cloth, and references to various weaving communities are common. Vijayanagara-period textile production has been studied by a number of scholars. Most important is the work of Vijaya Ramaswamy (1979, 1985a, 1985b, 1985c, 1993; also Mines 1984; Stephen 1997; Karashima 1992; Perlin 1983). South India’s textile trade was, of course, of great significance to Dutch, French, and British merchants and colonizers during the seventeenth and eighteenth centuries. A variety of colonial sources provide information on


textile manufacture during this period, when global mercantile and political forces contributed to dramatic changes in the organization and scale of production (Ramaswamy 1985c: 4). Much of what is recorded in these post-Vijayanagara sources is not directly relevant to the Vijayanagara period; nonetheless, descriptions of technologies, raw materials, and geographic distributions of weaving communities can provide valuable information for this study.

Weaving communities By the fourteenth century, South Indian textile production was highly specialized. Numerous castes of weavers are known; many have a long history that can be documented in pre-Vijayanagara texts and inscriptions. Weavers’ communities traditionally traced their origins to restricted geographic regions. However, many moved widely during the dynamic Vijayanagara period, leading to the expansion of some communities and contraction of others, as well as to increasing regional differentiation among widely dispersed communities of a single caste. Weaving communities in the northern regions of the empire – modern Karnataka and Andhra Pradesh – included the Devanga, Sale (also called Saliya or Saliga), Jeda (Jedara), and Senigar (Saniyar) castes. To the south, in the Tamil-speaking region, the Kaikkolar came to be the most important weaving community of the Vijayanagara period, displacing the previously dominant Tamil Saliya community in prominence (Ramaswamy 1985c: 15; pp. 187–190, this text). Large numbers of Kannada- and Telugu-speaking Devanga weavers were present in Tamil regions in the late eighteenth century (particularly in Congu District) and their ancestors are believed to have migrated there (and into Kerala) during the Vijayanagara period (Ramaswamy 1985c: 13). Other weavers included the Curubaru, a shepherd community that wove coarse woolen blankets, and the Pattunulkarrar, specialists in tie-dye fabrics. Along with communities who wove cloth, several other specialized occupational groups were associated with textile manufacture. Indeed, the Vijayanagara period appears to be marked by a dramatic increase in the degree of occupational specialization associated with textile production as compared to earlier periods. Inscriptions refer to dyers, fabric bleachers, tailors, and even cotton carders as named groups who owed taxes to local and imperial authorities. Spinners are not mentioned in inscriptions, though this task may have been similarly restricted to specific occupational groups particularly in areas of large-scale textile production. However, it is likely that spinning was

173

174


not a caste-specific or hereditary occupation, but was instead performed by a wide range of individuals of low economic and social status in cottonand silk-producing areas. Spinning appears to have been predominantly a female occupation, and a fifteenth-century poem by Srinatha refers to women spinning while men plowed the fields (Ramaswamy 1985c: 69). Sources bracketing the Vijayanagara period suggest long-term continuity in the organization of this task. In the early 1800s, Francis Buchanan (1988 [1807], vol. II: 263) observed that spinning in the Coimbatore region of Tamil Nadu (long a center of textile production) was carried out by “women of all farmers and low casts [sic].” He noted similar patterns in other areas he visited (vol. I: 218). And, from long before the Vijayanagara period, Ramaswamy (1985c: 2) cites a Tamil Çankam poem (late centuries BC to early centuries AD) that indicates that spinning was an occupation of widows. Such a pattern follows Dean Arnold’s (1985) model of specialization from below, in which individuals in difficult economic situations engage in specialized activities to provide supplementary income to subsistence agriculture or other productive activities. Cloth merchants were also widely involved in raw material (raw cotton and spun cotton and silk) and textile distribution during the Vijayanagara period. Some merchants were from weaving castes – and appear in inscriptions with the name setti (merchant) appended to their caste name. Others were from hereditary merchant communities. As discussed later in this chapter, inscriptional evidence suggests that, in some contexts, merchants may have had considerable control over production technologies and cloth quality.

Geography of textile manufacture From the fourteenth century until the present, cotton and silk textiles have been produced at a number of regional centers throughout South India (Figure 6.4). Many of these centers show remarkable continuity over the centuries, though a number of new weaving centers emerged during the Vijayanagara period, particularly in the Tamil-speaking Coromandel region of southeastern India (Stephen 1997: 73). Ramaswamy (1985c: 6) has recognized two major determinant parameters for the location of textile centers: (1) access to ports and transport routes; and (2) the availability of locally produced raw materials, particularly cotton and dyes. Cotton is grown throughout South India, primarily in dry farmed areas with deposits of moisture retentive clay-rich soils referred to as “black cotton soils.” Domingo Paes, recounting his travels from the west coast seaport of Bhatkal


175

ha d

hna Kris

T un

g ab

ra

VIJAYANAGARA Pen nar

Po

nn ai

Pallar ya r

Kaveri

Weaving centers

0

km

6.4 Centers of textile production in Vijayanagara-period South India (after Ramaswamy 1985c: 7)

through the territories of Vijayanagara, noted the presence of “an infinity of cotton” (Sewell 1900: 237). During the Vijayanagara period, major centers of textile production were found throughout Tamil Nadu and in inland Andhra Pradesh; other cotton-growing areas were found in upland Karnataka. There is no evidence for major concentrations of weavers at the first Vijayanagara capital, though a fifteenth-century text mentions that some of

300

176


Vijayanagara’s inhabitants were engaged in “colouring the cloths required during rituals” (Kotraiah n.d.j, XXXI, 15, p. 252). Weavers are seldom mentioned in summary lists of village servants (see pp. 100–102). However, it is likely that small numbers of weavers were present in many villages, where they worked as household specialists to provide garments to rural residents. Such weavers would have had a very different social and economic status than the weavers we know the most about – those clustered in and around temple towns. Differences between rural and urban weavers no doubt also existed in output, the quality of their products, and, perhaps, in the technology of production. It was the urban weavers of fine cloths who gained considerable economic and social prominence during the Vijayanagara period. Numerous new specialized weaving settlements appeared in this period. These took two forms: weaving villages, or weaving districts within temple towns. In either case, communities or districts were composed of weavers from a single caste. The geographic concentration of textile artisans implies a sophisticated distribution system – involving merchants, intermediaries, and a variety of market and transport systems. Subrahmanyam (1990c), following Brenning (1990), has presented Dutch census data from late seventeenth century Andhra Pradesh that provides information on the composition of some communities in the northern Coromandel region at that time. These communities postdated the Vijayanagara domination of the region and were linked to the Golconda kingdom. Nonetheless, in the absence of census data from the Vijayanagara period, the information provided may give a sense of the density and intensity of textile production and trade in resources and finished products in some regions of South India (see also chapter 7, discussion of the Alamkonda Puttasti). Data from 18 weaving centers in the east Godavari delta reveal an average of 331 weaver households per settlement, with approximately 418 looms (1.26 looms per household; Subrahmanyam 1990c: 94). The high density of textile workers that could exist in some communities is evidenced by census data from the small village of Gondawaram. In 1692, Gondawaram contained 91 households with a population of 431. In this admittedly extreme example, all but 3 of the households in the village were engaged in textile manufacture. Fourteen households of Kaikkolar weavers had a total of 15 looms (77 individuals: 20 adult men, 20 adult women, 37 children); 74 of the 91 households (81.3 percent; 337 inhabitants: 107 men, 113 women, 117 children) were “washers” whose primary occupation was washing woven cloth in the renowned alkaline water of the village reservoir. They also bleached, starched, and baled the washed cloth. These washers clearly served a broader community of weavers


than those residing in the village, and Dutch merchants were involved in transporting woven cloth to Gondawaram from considerable distances for processing (Subrahmanyam 1990c: 103–104). During the Vijayanagara period, weaving settlements were created through a variety of inducements. There is no convincing evidence for coerced movement of weavers, or indeed of any other craft producers, during this period. In contrast, there is good evidence that weavers were highly mobile and chose to move to enhance their social and economic conditions. Further, weavers and other artisans were able to move away from areas where conditions were not to their liking, most often to protest high tax rates. Inducements for settlement in weaving communities included tax reductions or abatements for specified periods of time, grants of agricultural lands, privileges in temple rituals, or positions in temple administration (see chapter 7). Weavers were thus at times attached to temples, but in a rather different way than many archaeologists have discussed attached specialization (see chapters 2 and 7; also Sinopoli 1998).

Textile technology The majority of textiles produced in South India during the Vijayanagara period were of cotton. Silk cloths were also produced, as were composite cotton cloths with silk edges; smaller quantities of linen and wool fabrics were also manufactured. A large number of named textile categories can be documented from inscriptions and texts (Ramaswamy 1985c). In some cases, textiles were named after the weaving community that produced them. For example, “Devanga” referred to a brocaded silk woven by members of the Devanga community, while Jedara Bhatigatu referred to a patterned silk produced by Jedara weavers. Other textiles were named after their place of production: “Kanchivani” refers to gold embroidered cloth manufactured in the temple town of Kanchipuram, which remains an important silk weaving center to this day. Along with these silk fabrics, a large number of named cotton cloth types are also described in texts, inscriptions, and travelers’ accounts. These were grouped into numerous categories based on fineness of weave, decoration, and color (Ramaswamy 1985c: 64). Market distribution of spun fibers suggests that ginning, carding, spinning, and perhaps dyeing largely occurred in areas where fibers were grown. While the spinning wheel was introduced into India in the fourteenth century (Ramaswamy 1985c: 20; and perhaps much earlier, Kenoyer, personal communications), it probably was not widely used in the South during the Vijayanagara period. Instead, spinners likely predominantly

177

178


engaged in hand spinning with spindles and whorls. Buchanan (1988 [1807]) described the spindles he observed in the early 1800’s as consisting of lumps of unfired clay attached to the ends of sticks, making this a highly portable industry, and one unlikely to preserve archaeologically. A probable spindle of low-fired (perhaps accidentally?) ceramic was recovered in an early or pre-Vijayanagara-period settlement site in the metropolitan region, and provides our only direct archaeological evidence of any stage of textile production. Spun fibers were widely available in local markets, were traded by mercantile associations, and were supplied to dependent weavers by “master weavers” and merchants (see below; Ramaswamy 1985b: 308). A variety of vegetable dyes were used to dye woven fabrics and yarns. These are mentioned in several inscriptions and texts. Safflower, madder, Sapan wood, turmeric, and lac (imported from Pegu) provided red colors, and indigo provided blue or black. Various mordants are also mentioned (Ramaswamy 1985c). Tie-dye and block printing were both used. Dyeing took place both before weaving (i.e., of yarns) or more frequently after weaving, when fabrics were taken to dye workshops to be dipped, painted, or block printed (Stephen 1997: 76). There is little information on loom technology during this period, though it is likely that several kinds of vertical and horizontal looms were in use (Ramaswamy 1985c). Loom technology and weaving techniques may have varied between weaving communities. An inscription from Tirupati, dated October 9,1538 (during the reign of Vijayanagara emperor Achyutadevaraya), documents an agreement between merchants and weavers concerning a particular style of weaving. It reads in part: . . . the following agreement is made between the cloth and yarn merchants of Tondamandalam, Purumandalam and Ulmandalam on one side and the lease-holders of Vijayanagaram, Magadhapattanam and Vidurapattanam and Purannappur: While we are assembling here . . . in the course of weaving by handlooms onethird of the Sadisarrakkuvadam . . . should be drawn lengthwise and 2/3 of the cotton should be used in cross-wise weaving; this mode of weaving should be done only by Muslims, and not by the Hindus. As a reward for their services in this style of weaving they are authorized to collect the incomes from the gifted lands for their weaving . . . We are all agreed by common consent that a fine of 12 gold varahan [coins] will be imposed upon anyone violating this rule, and paid into the temple treasury of Sri Venkatesa. We shall communicate this arrangement to every Hindu village, and Muslim dwelling, every cloth merchant and agent for strict observation applicable in

Craft products and craft producers Tirupati, Conjeevaram [Kanchipuram], and other parts of the south. We solemnly declare that we will not work in contradiction to this agreement. (TTDES IV,

GT148, pp. 208–210)

I will return to some of the non-technological implications of this unique inscription in chapter 7. In terms of the technological information provided in this inscription, Ramaswamy (1985c: 66–67) has suggested that the reference to Sadisarrakkuvadam, which she translates as “a four-cornered frame,” may refer to a horizontal draw loom. If so, this was a relatively recent introduction into South India in the fifteenth or sixteenth century (brought in from the north, perhaps through the expansion of Islamic artisans into South Asia), and was apparently restricted to limited groups of Muslim weavers. This inscription provides further support that textile production was characterized by a very high degree of technological specialization.

Textile use Throughout the Vijayanagara period, South Indian textiles were important in international commerce – with Europe and East Asia and with other South Asian regions. They were also important within South India, used to adorn structures and sacred images as well as for garments. Changing styles and demands for elite garments during the Vijayanagara period (discussed below) appear to have played important roles in the expansion of textile production, suggesting that both internal and external social and economic dynamics must be considered to understand productive intensification. Numerous literary sources of the Vijayanagara period provide information on both the opulence and the importance of textiles in urban contexts. Palaces and temples are described as adorned with flowing banners and elaborate textiles. For example, the 1648 text of Govinda Vaidya (the Kanthirava Narasaraja Vijaya; Kotraiah n.d.c) describes the royal audience hall at Srirangapattana during the annual Mahanavami festival (which closely paralleled the same festival that Paes described a century earlier at Vijayanagara, Sewell 1900): The audience hall was embellished variously. The main doors were tied with festoons and artificial screens fitted with gems, flags, banners, umbrellas, pinnacles on the top, festoons of tender leaves, mirrors, statuettes, all set with precious stones of different colors to make the whole show quite grand. (XX, 19–26, pp. 382–3; translated by

Kotraiah, n.d.c: 3.421)

179

180


Similarly, the Sivatatvacintamani composed by Lakkandadandesa, a military leader and provincial governor at the court of Devaraya II, described a royal ceremony at the imperial capital of Vijayanagara: the entrance doors of the palace, the shops in the city, the houses of the concubines, the residences of the officers of high ranks, homes of religious heads (and others) were all decorated in all possible grandeur using flags, banners, festoons, tassels, artistic works, kalasa [decorated water jars], mirrors (and others). (XXXI, 58,

p. 263, translated by Kotraiah, n.d.j: 3.421)

Sacred chariots used in religious festivals were also elaborately adorned with Gold, gems, colorful and white cloth bands, banners and tassels . . . clusters of coral, pearls, gems, and other precious stones . . . Further, long and wide cloth of different colors was hung from the top as part of the decoration. (Kanthirava Narasaraja

Vijaya, XXIII, 65–73, pp. 445–446; translated by Kotraiah n.d.c: 3.47)

Virtually all of the courtly texts of the period describe the use of textiles and other valued materials for adorning urban spaces in similar ways and provide evidence for large-scale consumption and display of elaborate cloths in urban settings. Vijayanagara texts and sculptural representations also provide abundant detail on the opulence and styles of elite garments. However, our best depictions of Vijayanagara clothing come from a sixteenth-century Shaivite temple complex in Lepakshi, Anantapur District, Andhra Pradesh (Paschner 1985; Gopala Rao 1969). Local belief attributes the construction of the temple to two brothers – Virupanna and Viranna – who served in the court of Achyutadevaraya. Virupanna appears as a temple donor in several inscriptions that date from 1530 to 1548, though none refer to the temple’s construction (Gopala Rao 1969: 45–46). The ceilings of the main temple and several other halls in the temple complex are densely covered with paintings; these are among the few extant paintings securely attributable to the sixteenth century and the Vijayanagara period. Although damaged, they give testimony to a rich and widespread craft that has left very little material evidence for scholars. The Lepakshi paintings recount a number of religious and historical tales, including events from the Mahabharata, Ramayana, and life of Shiva. While our limited sample of Vijayanagaraperiod paintings does not permit a detailed discussion of that craft, what is most interesting about these paintings for present purposes is their elaborate and detailed depictions of textiles. Both men and women are dressed in ornate and extraordinarily diverse patterned cloths. A portrait of two men,


6.5 Lepakshi painting: group of women attending the goddess Parvati (American Institute of Indian Studies, image 705–13)

perhaps Virupanna and Viranna, on the ceiling of one hall shows them as donors, with folded hands and wearing elaborate high-pointed kullayi hats and tailored white tunics. Around their waists are draped colorful patterned cloths. Women’s saris are draped in a variety of ways, which, along with their diverse hairstyles, may indicate distinctive ethnic or regional backgrounds (Figures 6.5, 6.6). In addition to depictions of garments, each of the painted panels is bordered by intricate floral or geometric designs that seem to derive from textile patterns (Gopala Rao 1969: 76). Indeed, Paschner (1985: 332) has suggested that the painters of the Lepakshi murals may have been primarily textile painters. Various lines of evidence indicate that clothing styles and textiles were the focus of both increasing elaboration and increasing demand throughout the Vijayanagara period. Several factors contributed to these expanded demands. First, the widening numbers and wealth of elites – at imperial centers, temple towns and regional centers throughout the empire – magnified the number of potential consumers for elaborate garments. Second, the political competition discussed earlier may have contributed to the prominence of

181

182


6.6 Lepakshi painting: group of women (American Institute of Indian Studies, image 705–82)

textiles and garments in displaying elite status. Third, the potency of textiles for conveying and absorbing elite status was manifested in new ceremonial contexts, which Gordon (1996, 2001) has termed “robes of honor” ceremonies. And fourth, interaction with Deccani Islamic polities to the north of Vijayanagara and the creation of new ideologies of kingship and display also contributed to new styles of elite dress (Wagoner 1996; Sinopoli 2000), which in turn also increased demands for certain kinds of textiles.


In a chronicle written between 1535 and 1537, the Portuguese horse trader Fernao Nuniz confirmed the potency of royal garments in his description of the Vijayanagara king Achyutadevaraya (1530–1542): The King never puts on any garment more than once, and when he takes it off he at once delivers it to certain officers who have charge of this duty, and they render an account; and these garments are never given to anyone. This is considered to show great state. His clothes are silk cloths of very fine material and worked with gold, which are worth each one ten pardaos [gold coins]; and they wear at times bajuris of the same sort, which are like shirts with a skirt; and on the head they wear caps of brocade which they call culaes . . . When he lifts it from his head he never again puts it on. (quoted in Sewell 1900: 383)

While the king’s garments were safely secured away after one use, unworn garments were important gifts in the Vijayanagara court, handed out to foreign ambassadors and subsidiary nobility. Domingo Paes, another Portuguese merchant resident in the court of Krishnadevaraya, reported that the king gave the departing Portuguese ambassador, Christovâo de Figueiredo, a “cabaya (tunic) of brocade, with a cap of the same fashion as the king wore” (quoted in Sewell 1900: 252). This ceremonial bestowal of royal robes is documented in a number of Vijayanagara texts; recipients included poets and religious authorities, ambassadors, nobility, and rulers of defeated states who received robes during ceremonies of “royal submission” (Gordon 1996: 232). These two styles of garments worn by Vijayanagara nobility – tunics and high conical caps – were adopted in South India during the Vijayanagara period and differed dramatically from earlier wrapped styles of royal dress (Wagoner 1996: 860). As Wagoner has documented, both their names and forms suggest that their origins can be traced to Islamic dress styles. The conical caps, also depicted in Vijayanagara paintings and sculptures (Figure 6.7), are called kullayi in Telugu (culaes by Nuniz). This word, and the form itself, derives from the Persian kulah, a type of high cloth cap known from Persian paintings from at least the fifteenth century (Wagoner 1996: 861). The ceremonial bestowal of royal robes had its origins in the courts of the Islamic polities of northern India, and is not known to have occurred in South India before the Vijayanagara period. Gordon (1996) has traced the origins and spread of the “robes of honor ceremony” in South Asia from the eleventh-century sultanates through the Mughal empire. Such ceremonies occurred in royal courts before an audience of elites and involved the presentation of garments “similar to those worn by nobles and the king” to the honoree, who “then took his place among the nobles in the court”

183

184


6.7 Sculpted figure on gate in outer city walls depicting pointed kullayi headgear


(Gordon 1996: 227). The bestowal and acceptance of ceremonial robes was a political act, simultaneously recognizing the authority and legitimacy of the donor, and the dependence and prestige of the receiver. Literary sources and travelers’ accounts document that the robes of honor ceremony was common in the Vijayanagara court, where it served the same function of representing and acknowledging political authority that it did in contemporary Islamic courts. The transformation of elite clothing styles during the Vijayanagara period apparently began among the royal family and was well established by the fifteenth century. By that time, tunics and peaked caps, such as those depicted in the later Lepakshi murals, had been widely adopted by elites at the imperial capital and by imperial tributaries in a process of emulation (Wagoner 1996: 859; the tunics were exclusively male garments, but the caps were reportedly worn by both men and women). In contrast, non-elites did not adopt these garments linked with royal identity and the royal body. Instead, throughout the period, commoners continued to dress in simple wrapped cotton cloths, saris for women and knee-length loincloths for men.

Weaving scale and concentration The intensification of textile production during the Vijayanagara period led to increases in overall numbers of artisans involved in these activities and to at least some changes in the organization of textile production. As noted, large numbers of weavers resided in close proximity in temple towns and urban contexts, particularly in inland regions of Tamil Nadu and southern Andhra Pradesh. Weavers of the same hereditary community or subcaste formed corporate units that regulated conduct, resolved internal disputes, cooperated in temple donations, purchased or leased land, and negotiated tax rates with local and imperial authorities (Ramaswamy 1985c). In some cases, communities of weavers were taxed as a single unit; in others, taxes were levied on individual looms (see chapter 7). Despite this evidence for corporate structures and increased numbers of textile producers, neither weaving technology nor the scale and organization of production units seem to have changed dramatically. Household workshops continued to be the major productive units for all aspects of textile production; and the vast majority of weaver households likely had only one loom and sold their products in local markets. However, we do find some evidence for the emergence of a new form of production, organized under “master weavers” or textile merchants (Ramaswamy 1985b). These were individuals who owned multiple looms that, in at least some instances,

185

186


may have been concentrated in a single physical structure or workshop (Ramaswamy 1985c: 84). Many of the individuals who controlled multiple looms were merchants. For example, information from Alamkonda from the early 1500s names two merchants who respectively owned 100 and 65 looms, and a third individual, Virayya, who owned 16 (Ramaswamy 1985b: 312; Nilakanta Sastri and Venkataramanayya 1939: 88–93; also chapter 7, this text). Even here, the majority of looms cited (235 of 411 looms weaving colored cloth) in this record appear to have belonged to individual singleloom households. Other owners of multiple looms were from weaving communities, and inscriptional evidence suggests that individual weavers were sometimes able to accumulate substantial wealth and influence. Two inscriptions that refer to an individual weaver are known from temples in nearby settlements of Thanjavur District (Tamil Nadu). This weaver is referred to by the term Kaikkolar mudali or “weaver chief,” suggesting that he may have been the owner of multiple looms or the head of a “workers’ group.” The first of these inscriptions dates to AD 1489 (ARIE 1940–378) and records the granting of a tax remission to one of the temple’s villages by the local Telugu-speaking ruler, Konerideva Nayaka. The inscription records that since this remission was granted at the request of the weaver chief, the ruler required the temple to grant that individual a house and temple lands. This property was specified to be inheritable, alienable, and tax-free (Karashima 1992: 75). The second inscription referring to this same weaver comes from another temple, and dates three years later (AD 1492; ARIE 1920–51). It commemorates a “voluntary” donation made by him to the same local ruler. This donation was used to sponsor a large ceremony (mahapuja) at the temple and in the construction of a new processional image. In consequence, the temple authorities granted the weaver a unit of temple land tax-free (Karashima 1992: 76). While a small number of “master weavers” are known from before the Vijayanagara period, they appear to have increased in numbers and importance after AD 1400, as a direct response to the increasing intensity of production and demand for textiles. We do not know what the relations between these individuals and the weavers who worked their looms were – i.e., their role in supplying raw materials or determining finished products – nor do we know how weavers who worked for them were recruited or compensated. However, Ramaswamy has suggested that the master weavers/owners may have had relatively little impact on textile forms. That is, there is no evidence for increased textile standardization resulting from the intensification of production, a pattern that most archaeological studies of craft production would lead us to expect (see chapter 2). Ramaswamy (1985b: 313) cites the


complaints of an eighteenth-century South Indian merchant who traded with the British East India Company. This merchant lamented that it was impossible to obtain identical pieces from a single weaver much less standardized products from multiple weavers. That merchants sometimes could exert control over textile technologies is, however, evident from the Tirupati inscription cited above, in which one form of weaving was agreed to be restricted to Muslim weavers. Vijayanagara rulers may have encouraged the emergence of multipleloom workshops. An inscription from Anantapur District (Andhra Pradesh) dating to the reign of Krishnadevaraya addressed tax rates for a new community of weavers in the settlement of Valpamadugu. It noted that no loom taxes were to be levied for the first three years of settlement, and after that, if families maintained ten looms, only nine were to be taxed (Ramaswamy 1985c: 84–85). I address taxation in more detail in chapter 7. Despite the evidence presented above, it must be acknowledged that concentration and intensification of production under master-weavers was not widespread, and household production continued to dominate. Both the 1692 census data from Andhra Pradesh and the sixteenth-century Alamkonda tax records cited above indicate that single-loom households were by far the most common mode of production throughout the Vijayanagara period.

The Kaikkolar weavers Of all the weaving communities of the Vijayanagara period, the Kaikkolar of Tamilnadu and southern Andhra Pradesh appear to have undergone the most dramatic changes in social, economic, and occupational status during the Vijayanagara period. They were, however, not unique in experiencing significant improvements in their status throughout the period; instead, they are representative of some general patterns of changing valuation of certain categories of craft products and producers, particular those that produced goods in high demand. I expand upon this point in chapter 7; here I limit my discussion to the Kaikkolar weavers. The Kaikkolar were a geographically extensive community, found over much of Tamil Nadu and southern Andhra Pradesh. At its broadest level, the identifier Kaikkolar references the jati or caste group, whose members shared a common hereditary occupation (weaving) and origin story (Ramaswamy 1985a: 429), and worshipped the goddess Kamakshi Amman (Ramaswamy 1979: 131). The name Kaikkolar, which appears in inscriptions, thus refers both to caste identity and to occupation. The Kaikkolar were also divided

187

188


into several gotra or lineages, and into hierarchically organized geographic or territorial divisions: four desai nadus, eighteen kilai nadus, and 72 nadus or localities (Ramaswamy 1979: 131). Each of these levels acted as corporate groups in various contexts: inscriptions that mention the Kaikkolar sometimes refer to actions affecting the entire jati, but more often refer to acts, such as the awarding of temple privileges, that affected Kaikkolar of one of the smaller geographic divisions or subgroups. It is not always easy to distinguish among these various referents, and here I speak of the Kaikkolar community quite broadly. The Kaikkolar first appears in the pre-Vijayanagara Chola-period inscriptional record in the eleventh century AD (Mines 1984: 12). Evidence from this period indicates that the Kaikkolar were already weavers and merchants and that they maintained armies, perhaps to guard regional trading ventures (Mines 1984: 12). Throughout the Chola period, the trading and military roles of the Kaikkolar seem to have been the predominant. Kaikkolar were members of the Ayyavole 500 regional trading corporation comprised of numerous merchant and artisan groups (Abraham 1988) discussed in chapter 4. Kaikkolar armies are mentioned in several twelfth-century inscriptions (Mines 1984: 13), and individual Kaikkolar are frequently referred to as members of the Chola emperor’s royal bodyguard (Ramaswamy 1985c). Kaikkolar origin stories also emphasize their military heritage. A twelfthcentury epic poem by the legendary Tamil poet Ottakuttar lauds the bravery and prowess of Kaikkolar soldiers, and traces their roots to the armies of the gods and the original Chola kingdoms of the Early Historic period (Ramaswamy 1985c: 14; Mines 1984). Along with their military occupation, many Kaikkolar men and women labored as weavers even during pre-Vijayanagara times. During the Vijayanagara period, evidence for Kaikkolar military activities disappears from the inscriptional record, perhaps as a result of imperial or regional leaders’ attempts to suppress local military forces. While their military role declined, the Kaikkolar as weavers rose to great prominence in the Tamil region. As noted earlier, they appear to have largely displaced the previously dominant Saliya weaving community in this area. The social and political dynamics underlying these changes are not well understood, although their pattern is clear. Streets of Kaikkolar weavers, called perunderuvu (“large streets of many looms”) are documented at numerous temple centers throughout the region, as are several new Kaikkolar villages (Stephen 1997: 73–74). The enhanced social status of the Kaikkolar community throughout this period is perhaps best evident in their expanded role in temple


Table 6.1 Kaikkolar weavers in Tirupati-Tirumalai inscriptions

Dates

References to Kaikkolar +

Total dated inscriptions referring to craft producers∗

Total # of inscriptions (with year dates)

pre-1450 AD AD 1451–1500 AD 1501–1565 AD 1566 – 1700 Totals

1 1 24 11 37

2 4 36 14 56

55 (36) 166 (122) 653 (435) 192 (34) 1066 (627)

+does not include 2 undated 15th-century inscriptions ∗ does not include 3 inscriptions which can not be precisely dated

administration and in the sheer frequency of references to them in inscriptions and other written sources. An examination of inscriptions from Tirupati and Tirumala temples in southeastern Andhra Pradesh serves to illustrate these changes (Table 6.1). Craft producers are mentioned in fewer than 6 percent of the Tirumalai-Tirupati Vijayanagara-period inscriptions, only 59 out of 1066 (5.3 percent), far less frequently than kings, nayakas, merchants, or agriculturalists. Of these 59 inscriptions, however, Kaikkolar appear in two-thirds (n=39), many times more often than any other craft group. As discussed in chapter 7, they appear most often in lists of communities who were to receive offerings generated from donations made to temples by a variety of donors. When artisans such as potters are mentioned as recipients of offerings, it is in return for vessels they provided to the temple; that is, for performing their service as potters. In marked contrast, Kaikkolar received distributions for their services as temple officers or functionaries, and not for providing textiles. Among the tasks entrusted to Kaikkolar were the rights to decorate temple structures for specific ceremonies and the rights to carry sacred images or vehicles through the streets of Tirupati during processions. The remuneration for these services was substantial, and ranged from one to three panam per service, as well as a share in the prasadam (sacred food offerings from the temple; Ramaswamy 1985c: 107). Other local communities of Kaikkolar weavers also were granted privileges during the Vijayanagara period. An inscription from the Tamil temple center of Tiruvannamalai, dated to AD 1414, records the granting of privileges to the local Kaikkolar community by the Vijayanagara king. These included rights to ride on palanquins and elephants, to carry flywhisks, and

189

190


to blow the sacred conch, all marks of considerable prestige and ritual standing (Srinivasan and Reiniche 1990; also SII/71/155). Similar privileges were granted to the Kaikkolar weavers of Kanchipuram and surrounding districts (Ramaswamy 1985c: 105). As discussed earlier, individual Kaikkolar, described as weaver chiefs, were able to accumulate significant personal wealth, including agricultural land and other property, and were in a position to make donations to temples in their own names. Groups of weavers are also sometimes documented as having made collective donations, alone or in concert with other artisans, such as temple women. I will return to Kaikkolar–temple relations (and weavers and temples more broadly) in chapter 7.

Metal workers South Indian metal workers of the fourteenth through seventeenth centuries predominantly belonged to a distinctive fivefold hereditary community that included blacksmiths, copper or brass smiths, goldsmiths, masons and sculptors, and carpenters. This grouping can be traced back at least to the early Chola period (eighth to ninth century AD). Its members were called “Kammalar” in Tamil, “Panchalar” in Kannada, and “Panchanamuvaru” in Telugu. They were also referred to as “Rathakarar”, “Asari”, or “Acharya” (Ramaswamy 1993: 27–28). Ramaswamy (1993) has argued that the origins of this distinctive multioccupational grouping can be traced to the shared labor of these artisans in temple construction. She noted that references to the community first appear in the Chola period, when investment in the construction of temples and temple towns underwent a significant expansion. The term Rathakarar (or “chariot makers”) is, Ramaswamy suggests, a clear reference to temple activities, specifically the construction of processional chariots. The work of these diverse artisans in the various tasks of temple construction and maintenance may have been an impetus to their development of a corporate identity that persisted through the Vijayanagara period and, indeed, continues to the present. An occupational identity that began in temple contexts eventually came to encompass producers in non-temple village and urban contexts as well. Today, this fivefold community calls itself the “Visvakarma”, and in oral and written traditions traces its origins to the deity Visvakarman. This god appears in the Rig Veda, the most ancient of India’s sacred texts, as the architect of the Gods (Brouwer 1995). For the Visvakarma, their patron deity


(also called Virat Visvabrahma) is, however, far more than a laborer for other gods. Instead, he is the “universal essence” and source of all creation, and they are his direct descendants. An 1818 Visvakarma publication recounts their origin story: The body of the Virat Visvabrahma consists of the created Universes only. This world (earth) which is called Brahma is the feet of that Virat Visvabrahma, Vishnu its navel (corresponding to water), Mahesvara its midriff (corresponding to air), Sadasiva is the sky, which forms the head of this (created) earth. When on earth, there was neither fire, nor air, nor light, nor consciousness, nor stars, there was only the self-contained Virat Visvabrahma. The Brahma of the five faces emanated from that form. The five faces . . . gave rise to the panchabrahmas. (quoted in Brouwer

1995: 42)

The panchabrahmas, representing “the five pillars supporting society” (Brouwer 1995: 43), were in turn the progenitors of producers of the five crafts – blacksmiths, coppersmiths, sculptors, goldsmiths, and carpenters. Given their claimed origin from the universal source from which the other gods emerged, the Visvakarma do not acknowledge the ritual superiority of Brahmans and Kshatriyas. Instead, they claim a status apart from the conventional caste hierarchy. In contemporary contexts, these claims are not widely accepted by non-Visvakarma, who consider this community to belong to the Shudra varna, with low social and ritual status. While we do not know precisely when this particular origin story developed, it certainly existed by the late eighteenth century, and may extend back far earlier. The increases in the status of smiths and other artisans that occurred during the Vijayanagara period may have contributed to the development of this distinctive assertive identity. Members of the fivefold smith community share a unitary identity at a broad regional level and sometimes appear as a single category in Vijayanagara-period inscriptions (as Kammalar or Panchala). However, each of the distinctive occupations within the group is also recognized, and individuals are defined as masons, gold or iron smiths, carpenters, sculptors, etc. The extent to which occupations were fixed within the broader Kammalar community is difficult to discern. Many (most?) individuals appear to have engaged only in a restricted range of craft activities; others may have moved between these various activities or occupations. There is inscriptional evidence that some individuals did not follow the occupations of their parents: for example, an inscription engraved on a reservoir embankment in Shimoga District in southern Karnataka in AD 1526 is signed by Viroja, who calls himself “an engraver” and notes that he is the son of the

191

192


goldsmith Gandroja (EC VII–VIII, supplement, p. 324). However, the extent of occupational mobility within the Kammalar caste is unknown, and it is evident that the many distinctive occupational subgroups that comprised the larger category retained distinctive identities, and defined themselves on the basis of occupation, language, and place of residence. As a community, the Kammalar underwent similar economic and social enhancements during the Vijayanagara period to those experienced by Kaikkolar weavers. Nonetheless, such changes were not uniform or equally distributed, and considerable variation existed within and among various Kammalar artisans. The lives, statuses, and privileges of village blacksmiths and urban master goldsmiths were quite different. I will return to a discussion of some of the broader patterns of Kammalar experiences during the Vijayanagara period in chapter 7. First, I briefly consider archaeological and textual evidence for the individual metal-working crafts produced by members of the Kammalar community. I limit my discussion to those metal crafts for which we have good archaeological or textual evidence. I therefore do not discuss the working of precious metals. Gold and silver ornaments and vessels were no doubt produced in large numbers during the Vijayanagara period, and are mentioned in numerous texts. However, relatively few well-provenienced Vijayanagara precious metal objects have been documented. Many such artifacts likely exist in private collections and in museum and temples storehouses; many more were no doubt recycled and reworked. However, other than occasional references to finished objects or goldsmiths in texts and inscriptions there is little evidence that can be used to discuss technologies or structures of production of precious metals and stones.

Iron working Numerous sources of high-quality iron ore occur throughout South India. Closest to Vijayanagara is the area known as the Sandur Hills, located on the southern edge of the Vijayanagara metropolitan region. Iron working has a very long history in the region, where it can be traced to at least the early first millennium BC (Chakrabarti 1992); evidence for the production of crucible steel dates to the early centuries AD (Rajan 1997; Craddock 1995: 279, and below). This South Indian steel or wootz was in high demand in international commerce; from at least the third through the eighteenth centuries AD, steel ingots were shipped to the Middle East and used in the production of high quality swords of “Damascus steel.” Steel and iron were shipped from the seaport of Bhatkal in western India in the fifteenth and early sixteenth


centuries (Ramaswamy 1985a; Subrahmanyam 1990b), and from the sultanate of Golconda (Andhra Pradesh) in the sixteenth and early seventeenth centuries (Methwold in Moreland 1931: 37–38). During the Vijayanagara period, large-scale iron and steel production appears to have been focused in source areas in inland Andhra Pradesh and western Karnataka (Ramaswamy 1985a: 424–425; Subrahmanyam 1990b: 30; this area has been the focus of recent, largely unpublished, research by Thelma Lowe). In addition to these centers of large-scale production, blacksmiths and smaller-scale smelting facilities were widespread in rural and urban contexts throughout South India – producing and repairing cart wheels and agricultural tools, and manufacturing swords, arrow points, guns, armor, and other military hardware. Smiths frequently appear in lists of village servants or ayagar (see chapter 4). They are also mentioned in many Vijayanagara-period texts in lists of artisans who accompanied military troops and worked in military encampments. For example, Nanjunada Kava’s Ramanatha Charite (c. 1525 AD; Kotraiah n.d.h), a biography of the hero Ramanatha of the fourteenth-century Kampili kingdom, describes the artisans of the ruler’s military camp. Blacksmiths are reported as having made and repaired axes, swords, bars, and other weapons. Other metal workers mentioned in texts included tool sharpeners, who honed swords and daggers, and tinkers who made and repaired armor, helmets, face-guards, and the like. Similarly, Virupaksha Pandita’s description of military camps in the Channabasava Purana (AD 1585) mentions blacksmiths engaged in the production and repair of “steel-axes, pick-axes, crow bars, sickles, spikes, swords, daggers, [and] other cutting weapons” (Kotraiah n.d.b: XLVIII.42).

The technology of iron smelting and forging Although we have no written descriptions of the smelting process from the Vijayanagara period, Francis Buchanan (e.g., 1988 [1807] vol. I: 170–177; vol. II: 16–21) and other eighteenth-century travelers left detailed descriptions of the smelters and smiths they observed during their travels through South India. Crucible steel production has been documented most thoroughly, given the great interest of Europeans in this sophisticated technology. The workshops described by Buchanan varied in the number of workers employed, seasonality of production, and output. However, all were relatively small, employing eight to fifteen workers under the supervision of a senior smith. Buchanan does not describe who the workers were or how the production unit was constituted. Some wage laborers also seem to have been employed in iron production and mining. Along with the laborers

193

194


who processed and smelted ore, specialists associated with iron production included charcoal producers and miners. In the workshops that Buchanan observed, the same individuals who engaged in smelting also mined or gathered ore; this may also have been the dominant pattern of the Vijayanagara period, though was perhaps not the case in larger-scale production facilities. Iron smelting consumes large quantities of ore and charcoal, and access to both is a major determinant of the location of smelting sites. As noted above, South India is rich in iron sources and, during Vijayanagara times, forests still existed in many areas. However, iron production may well have played a significant role in increasing the rate of deforestation in some regions. Other raw materials necessary for iron smelting are more readily accessible: sandy clay to construct furnaces, coarse tempering materials (such as quartzite) to enable them to withstand the high temperatures necessary for smelting (above 1100◦ C), and water. Additional tools include hide bellows, and hammers and stone anvils for forging. South Indian iron furnaces as known from both archaeological and historic sources were typically relatively small single-use facilities (thus many furnaces may have been constructed at a single site over a period of weeks or months). Production output per smelting episode was also relatively low. Individual episodes observed by Buchanan yielded between five and nine kilograms of useable iron after smelting and forging. Output varied depending on quantity and quality of the ore used; typically iron ores contain high percentages of silica, carbonates, and minerals that must be removed through smelting to produce workable iron. The furnaces used in smelting were small cylindrical shaft furnaces, with heavily tempered clay walls. Diameters typically range from c. 15 to 40 cm (Craddock 1995: 170–171). The difficulties in attaining the high temperatures necessary to melt iron (1550◦ C) precluded casting. Instead, ores were smelted. The smelting process entails heating ore to remove oxygen and impurities (as slag) to produce a relatively pure spongy iron mass, known as bloom, which is then worked through smithing (repeated heating and hammering to remove remaining waste). In smelting, ground ore (or iron-rich sand) is placed inside the furnace along with large quantities of charcoal. Bellows force air into the furnace, enabling the fires to achieve high enough temperatures to remove the impurities in the ore under the reducing atmosphere necessary to remove the oxygen. A smelting event observed by Buchanan (1988 [1807], vol. I: 170–177) in the town of Magadi, west of Bangalore, is representative of smelting techniques throughout the region. The facility employed four charcoal makers, four miners (who worked four months a year), and four smelters. Two


additional charcoal workers worked in the forge, as did six “laborers.” A single smelting consumed approximately 34 kg of iron-rich sand and three bushels of charcoal, and firing lasted approximately 6.5 hours. At the end of the firing, the furnace was broken apart to remove the bloom, which weighed 11 kg. The bloom still contained high levels of impurities and had to be repeatedly heated and hammered by forgers to produce a workable piece of iron (Craddock 1995: 235). Both the smelting and smithing process produced slag. The final product of useable iron weighed roughly 4 kg, approximately 12 percent of the weight of the ore. Buchanan noted that smelters from nearby regions reported that output was somewhat greater when working with ground ore, as opposed to the sand, yielding a proportion of approximately 20 percent workable iron to total ore consumed. Given the high ratios of waste slag to useable iron produced in smelting and forging, the debris produced at even relatively small smelting sites is considerable. While furnaces themselves are seldom preserved, smelting sites are readily identifiable by large quantities of iron slag, as well as the ceramic fragments of destroyed furnaces, and ceramic pipes or tuyeres, through which air is forced into the furnaces. In all of the workshops that Buchanan visited, smithing or the process of removing impurities from the bloom took place at or near the smelting facilities; blocks of workable forged iron were then distributed to blacksmiths or middlemen. The blacksmiths also produced some slag through additional working of impure iron, but the greatest amount of waste material is likely to occur at smelting locales.

Archaeological evidence of iron working in the Vijayanagara Metropolitan Region Although historical sources suggest large-scale iron production in several areas of the Vijayanagara empire, the archaeological evidence from the region of the first imperial capital indicates relatively small and mobile production units (see Lycett 1995). The primary category of archaeological evidence for iron working is slag, produced from both smelting and the smithing of bloom. Iron slag was recovered in a large number of sites in the Vijayanagara metropolitan region and in the Vijayanagara urban core (Sinopoli 1986; Morrison 1995). Low slag densities were recorded in the urban core and in more than thirty sites in the metropolitan region, characterized by isolated and dispersed small slag fragments. Low densities of slag also occur in non-site locales. These may be the result of smithing activities or of a more generalized redistribution of surface remains, as a result of manuring or other waste distribution practices. A few non-smelting sites have much

195

196


higher slag densities. Most striking is the high Vijayanagara-period platform, VMS-7; located approximately sixty meters to the west of smelting site VMS-121 (described below). The fill of this approximately 2.25-meter high platform consists largely of iron slag and other waste materials from the nearby production site. Seven sites in the metropolitan region can be classed as iron smelting and, possibly, forging sites (Table 6.2, Figure 6.8); one of these (VMS-110) dates to the early centuries AD and is not discussed in detail here. They are characterized by dense scatters of iron slag and droplets; fired ceramic furnace fragments, often with slag adhering; and the remains of ceramic tuyeres. Small pieces of very high-quality iron ore were identified in VMS121 and VMS-179 (Gogte n.d.: 6). These were cobbles, most likely collected in streambeds near the Sandur Hills, where they are abundant. Extensive mining was probably not necessary for these relatively small production contexts. Other than rare fragmentary remains, traces of furnaces are, not surprisingly, sparse. VMS-121 contains what may be the remains of such a facility in a low rectangular platform, c. 3 × 5 m. Fired earthen or ceramic walls and a circular ceramic tuyere fragment project from the platform surface, though it is not possible to determine whether these are in situ, as the area has been converted to a naga (sacred snake, associated with fertility ritual) shrine and is presently in worship (Figure 6.9). However, a small irrigation canal has recently been excavated about thirty meters to the south of the platform and exposed the remains of another small firing facility – a basin-shaped feature containing cinder and slag. Gogte (n.d.: 16) has suggested that smelting at this site may have been conducted in relatively simple hearths, rather than in more elaborate shaft furnaces. Five of the six iron-working locales (all except VMS-6) contain no formal architecture. Instead, they are located in flat open areas within a kilometer of defined settlement areas. They are all near water sources, particularly reservoir embankments. While some of these sites have quite extensive remains of waste materials – for example, VMS-5 with a slag scatter extending over approximately five hectares – the absence of residential architecture and other permanent facilities suggests that these were not long-term production sites with permanent resident artisans. Instead, the sites appear to be the result of the accumulation of multiple episodes of small-scale smelting and forging activities. Lycett (1995) has proposed that the association of many of these sites with seasonal water sources indicates that smelting activities occurred on a seasonal basis.


197

Table 6.2 Iron-working locales in the Vijayanagara metropolitan region Site

Dimensions

Description of remains

VMS-5

280 × 200 m

Dense scatter of iron slag, small droplets of iron waste and localized areas of vitrified ceramics suggesting multiple small-scale production locales, on the edges of the catchment of reservoir VMS-4.

VMS-6, Feature 2

135 × 130 m

Settlement locale associated with fortified outcrop. Contains slag, vitrified ceramic and pipe (tuyere) fragments, found throughout settlement area with highest densities in the northwest quadrant.

VMS-110

c. 200 × 200 m

Large mound of iron slag and associated scatter. The mound has been converted to a Vijayanagara-period reservoir. The iron-working debris is however, pre-Vijayanagara and dates to the early centuries AD (see Sinopoli and Morrison, in press; Morrison 1998).

VMS-121

150+ m diameter scatter

Small platform with remains of iron-smelting furnace. The platform is presently the location of a naga shrine, but beneath the shrine are projecting tuyere fragmens and over-fired ceramic furnace fragments. Other possible furnace remains are visible in a recent irrigation cut. Surrounding the platform is a dense surface scatter of iron slag, overfired ceramic, iron droplets and tuyere debris. Slag from this site has been used as fill in Vijayanagara-period platform, VMS-7, and a light scatter extends more than 500 m to the south.

VMS-169

150 × 150 m

Diffuse scatter of lithics, iron slag and ceraimcs. Slag densites are highest in the eastern half of the site. The site is located in an upland area between reservoir embankments VMS-165 and VMS-190.

VMS-179

50 × 35 m

High density of iron slag and vitrified ceramics in localized clusters.

VMS-317

60 × 50 m

Shaivite temple complex and associated mandapas. A moderate to high-density slag scatter is located to the south and east of the temple, associated with a mound of fired earth and clay, which may be the remains of a furnace.

As noted above and evident in Table 6.2, surface scatters of slag and other smelting waste around localized areas of smelting are often quite extensive. For example, low quantities of slag were observed as much as half a kilometer to the south of smelting site VMS-121; a similar light but extensive distribution was observed around VMS-5. It is likely that these extensive scatters result from the redistribution of ash as fertilizer for agricultural fields. That is, as for ceramics, we believe that manuring activities have led to considerable redeposition of organic waste from both household contexts and industrial facilities across the metropolitan region. Small pieces of slag

198

The Political Economy of Craft Production d d d

nd

iC

ana l

0 40

r ive aR a dr

Tu ng a

bh

o Aneg

Anegondi

na

l Turtha Cana

Kalaghatta C

l

i

a

Urban Core

VMS-179 i

VMS-317

d

Raya Canal Kamalapuram Reservoir

VMS-169 VMS-121

VMS-5

Vijayanagara period canals

0

km

Fortification walls areas systematically surveyed

6.8 Iron-working sites in the Vijayanagara metropolitan region

Contour interval 20 m Lowest contour 400 masl

5


6.9 VMS-121: Iron-working site and shrine

could easily have been incorporated into loads of ash, resulting in a sparse extensive scatter of the kind we observe. Site VMS-6 is an exception to the general distribution of iron-working sites in non-residential areas. This site consists of small fort constructed on top of a hill, and an associated settlement on the plain beneath it (Figure 6.10). The settlement, presently just under two hectares in area (though recent stone quarrying has likely reduced the original area considerably), contains dense, well-organized rubble wall architecture and the remains of domestic activities. The dating for the site is uncertain; based on architectural form, it either dates to the late Vijayanagara occupation in the region or may post-date the 1565 abandonment of Vijayanagara. Numerous large pieces of slag, over-fired ceramic pieces, and furnace fragments are present in the northwest quadrant of the settlement. The production of iron implements in a site that was apparently a military encampment suggests the presence of smiths and smelters associated with military activities, perhaps responsible for the production of weapons or other military gear. The problematic chronology of this site makes it difficult to determine precisely when this production occurred. In general, the archaeological pattern of iron working in the Vijayanagara metropolitan region provides evidence for small-scale producers who moved across the landscape seasonally to serve a variety of consumers. These

199

200


modern road

0

m

50

6.10 Hilltop fort and associated settlement (VMS-6): shaded area indicates mounded rubble area, dashed lines indicate projected wall alignments


workshops may have served agriculturalists in nearby settlements, or they may have been associated with other activities, such as military encampments (e.g., perhaps in the flat open area around VMS-121, near the associated large Vijayanagara platform), which we know from historical sources were located on the outskirts of the urban core. Again, larger-scale production centers likely existed in other areas of the empire, particularly for the production of iron and steel for military purposes and for export, or even in the nearby Sandur hills, where they have likely been destroyed by contemporary large-scale mining activities. Historically documented steel and iron products of the Vijayanagara period included large and small guns and cannons, cannon balls (large numbers of which were reportedly exported to Europe; Subrahmanyam 1990b), swords, spear points, etc. While some larger production sites have been documented, these have not yet been published, and we know little about how the production groups were organized, that is, whether there were large numbers of small production units, as we see among most weavers, or if large-scale units were present.

Crucible steel production Steel is an alloy of iron and carbon. It is much stronger than iron alone, and was particularly desired for weapons, such as swords. Given iron’s high melting temperature, the production of steel is a complex process, rather different from the smelting process described earlier. If not properly produced, steel becomes very brittle and is difficult or impossible to work (Craddock 1995: 236–237). As noted above, Indian steel was in considerable demand during the Vijayanagara period, both for external maritime trade and within South India. South Indian steel was produced using ceramic crucibles, in which smelted iron and carbon-rich organic materials were fused or “carburized.” Crucible steel (or wootz) production has a very long history in South India, going back to at least the early centuries AD and perhaps a few centuries earlier (Rajan 1995). High quality Indian steel was in great demand in the Middle East and the Mediterranean from the Roman period through the eighteenth century AD. Because of the great interest in Indian steel in the west (even into the colonial period), the process of crucible steel production is comparatively well documented. The earliest incontrovertible textual reference dates to the third century AD, and several earlier sources also appear to refer to this material. From more recent periods, we have several eighteenth- and nineteenth-century descriptions of Indian crucible steel production (summarized in Bronson 1986; Craddock 1995), and recently some metallurgical

201

202


and chemical analysis has been carried out on archaeological materials (e.g., Lowe 1989; Lowe, Merk, and Thomas 1991). Lowe has documented more than forty historic production sites in Nizamabad District (Andhra Pradesh), many of which date to the Vijayanagara period, though only preliminary reports have been published to date. Crucible steel production was carried out through sealing fragments of wrought iron and organic materials, such as leaves or wood, in ceramic crucibles. The crucibles themselves were quite small, and heavily tempered (to enhance their refractory qualities), with a total capacity of only a few hundred cubic centimeters (Craddock 1995: 276). These crucibles were then placed in a fire and heated to extremely high temperatures, usually for several hours, until the iron had absorbed sufficient carbon to be able to melt (the addition of carbon to iron lowers its melting temperature by several hundred degrees). The crucibles were then air-cooled or, in some cases, rapidly cooled through dousing; these different techniques producing somewhat different end products (Craddock 1995: 278, has suggested that gradual cooling was necessary to produce the most desired wootz). Buchanan described the process of crucible steel production at the Magadi workshop: The same persons [who smelt and forge iron] also make steel. Good clay is mixed with an equal quantity of the charcoal that is made from Paddy husks; and having been well moistened with water, is thoroughly mixed, by being trodden under the feet of oxen. It is then picked clean, and made into cuppels [sic], which are then dried one day in the shade and next day in the sun. A fire place is then made, in form of a parallelogram, by placing two stones one cubit long, and two inches and a half high, parallel to each other. At the distance of a foot above the stones is placed a wall of clay eight inches high. One end is shut up, in the same manner, by stone and clay; the other is built up with clay alone to the height of two cubits. Through this is inserted a tube for two bellows. Each of the cuppels is now loaded with small pieces of iron, from one to one and a half Seer (9 7/10 oz. to 14 oz.) in weight, together with five small pieces of the Tangayree wood (Cassia auriculata). Three rows of loaded cuppels are placed one above the other, so as to occupy the whole area of the furnace; the room of one cuppel only being left empty, opposite to the muzzle of the bellows, in order to give access to the wind. They are covered with two bushels of charcoal, and burned for six hours; a third bushel of charcoal having been added, as the former two were consumed. The pieces are then taken out, and hammered into small squares, having been heated with charcoal of the Sujalu (Mimosa Tuggula, Buch. MSS.). (Buchanan 1988 [1807], vol. I: 174–175)

Only one crucible has been recovered from the Vijayanagara metropolitan region and it appears to have been associated with copper working, rather than steel production, and likely dates to the pre-Vijayanagara period.


Steel ingots from South India were traded to the Middle East where they were forged into elaborate blades known as Damascus steel (Craddock 1995: 275). Although we have few Indian steel weapons from the early Vijayanagara period, elaborate seventeenth- and eighteenth-century steel weapons are known from South Indian Nayaka courts. The most important seventeenth-century collection comes from the armories of Thanjavur, Tamil Nadu (Michell (1995: 215–219). Weapons include swords (pattar), daggers (katar), and ceremonial elephant goads (ankusha). The gilded iron, copper, or ivory hilts of these finely crafted steel blades were quite ornate, with three-dimensional animals, including snakes, lions, elephants, peacocks, and mythical creatures, as well as scrollwork, foliage, and deities.

Copper and bronze working Copper and bronze images of deities and saints comprise part of the necessary ritual paraphernalia of South Indian temples. In temple sanctuaries, the central non-movable temple images are most often made of stone. However, bronze portable images are essential for processions and other rituals that involve viewing or worshipping the god outside of the temple sanctuary (Michell 1995). Large and even smaller temples can possess dozens of such images of deities, their vehicles, and associated saints. Many Shaivite temples contain bronze images of the sixty-three saints or Nyanmars, while bronze images of the twelve Vaishnava Alvar saints are also common (Michell 1995: 201). Images of Jain saints, Buddhist figures, and bronze images were also produced during the Vijayanagara period. Portrait images, of royal figures or elite donors, also came to be produced in increasing numbers during the fifteenth and sixteenth centuries. The most famous of these are the portraits of Krishnadevaraya and the two queens Chinnadevamma and Tirumaldevamma in the Venkateshvara temple at Tirumalai (Michell 1995: 204–206). Smaller, portable bronze images were likely also part of household worship, and bronze lamps, vessels, and other artifacts also were in use during the Vijayanagara period, in temples as well as in affluent households. Numerous workshops must have existed to produce the large numbers of bronze images and artifacts known from fourteenth- through seventeenthcentury South India, and it is likely that these workshops were found in urban centers and temple towns throughout the empire. However, Vijayanagara bronze production and even finished products have been little studied and no workshop sites have been documented. Art historians (e.g., Nagaswamy 1983, 1988, 1995; Gangoly 1978 [1915]) have largely focused their attention on earlier bronzes of the Pallava and Chola period, deemed to represent the

203

204


apogee of South Indian bronze working. Vijayanagara-period bronzes are typically viewed as being of much lower artistic quality than earlier works (Nagaswamy 1988: 178–179; Shetty 1988: 87). Even for the relatively wellstudied Chola bronzes, traditions or schools of casting believed to represent distinct workshops or casting traditions are attributed on the basis of stylistic evidence and provenience of sculpture recovery, rather than through identification of workshops. In fact, we cannot link any sculpture to a named artist or workshop. This is because the vast majority of bronze images were neither signed nor dated; instead, “schools” or traditions are named after prominent royal temple sponsors. Although much has been written about South Indian bronzes, scholarship has typically focused on the highest quality pieces, with little consideration of the range of variation between workshops and within periods. Chronological attributions are themselves problematic, since they are primarily made stylistically, through comparisons with the small number of well-dated images. However, Michell has argued (personal communications; also Srinivasan 1999) that Vijayanagara bronze sculptors continued to use earlier stylistic conventions in their work. He believes that many bronzes that have been attributed to the Chola period may in fact have been produced significantly later, and that conventional stylistic dates must be treated with caution. Archaeometallurgist Sharada Srinivasan (1999) has recently provided an alternate method for dating, on the basis of metallurgical content identified through trace element and lead isotope analysis of more than 130 South Indian bronzes in museum collections. Included among these were a number of inscribed and dated sculptures; the others were tentatively dated on stylistic grounds. The trace element studies focused on eighteen relatively rare elements whose presence could result either from the exploitation of discrete ore sources, recycling old bronzes, manufacturing processes, and/or material additions made by casters and smelters. Thus, while elements could be identified, their sources could not. Elemental compositional studies are made more complex by the increasing complexity of alloys produced beginning in the fourteenth century. Early images were predominantly composed of copper (more than 90 percent; Nagaswamy 1988: 144). In the fourteenth century, South Indian casters began to add small quantities of gold, silver, and zinc, to the tin and copper. The resulting alloy was called pancha loha – the “five metals” – and, according to textual sources, merged the sacred properties of each metal into a unitary whole. Thus, gold was linked to enjoyment in this world and the next and to victory, silver to wealth, copper to well-being, and so on (Nagaswamy 1988: 144). The five


metals also represented the five elements of earth, air, fire, water, and ether, and symbolized the all-encompassing nature and pervasiveness of the sacred (Nagaswamy 1983: 8). Because of the various additions made to ores, the presence of individual elements in bronzes is not useful for identifying ore sources. Other trace elements may be more relevant to source identification, but to date too few ore sources have been characterized to make this possible. Nonetheless, analysis of trace element distributions from even poorly provenienced bronzes can still be useful for identifying chemical groupings that can result from technological and/or temporal differences in material selection and preparation. Srinivasan’s trace element analysis suggests that it is possible to identify chronologically discrete chemical signatures based on element frequencies and ratios. For the Vijayanagara period, Srinivasan identified two distinct clusters. She termed the first “High Vijayanagara” (Srinivasan 1999: 106) bronzes; these were fine sculptures that were stylistically similar to imperial Chola bronzes, but with a distinctive chemical signature. The second group was termed “Lesser Vijayanagara bronzes” (Srinivasan 1999: 106), with stylistic parallels to provincial Chola sculptures. South Indian bronzes were produced using the lost wax casting technique. A wax model of the image was first produced. This was covered in clay to form a mold into which molten metal was poured. The majority of South Indian bronzes are solid, though a small number of hollow images were also made. Numerous centers of production existed in the Tamil region during the Chola period. These included urban sites and political capitals (e.g, Gangaikonacholapuram, Thanjavur) and major temple centers. The latter included Kumbakonam, Kanchipuram, and Tiruvannamalai during the early Chola period, and, later, Tanjore, Thiruvenkadu, and Thirukkarvasal, among others (Nagaswamy 1988: 167–169). Bronzes were also produced in Kerala, Andhra Pradesh, and Karnataka. As noted above, the identification of these centers is based on finds of finished sculptures and stylistic similarities, rather than direct archaeological evidence of workshops. Nonetheless, art historians do argue that workshop traditions can be identified. Some of these prominent workshops appear to have endured for multiple generations. Several Chola-period temple inscriptions record the commissioning of bronze sculptures, and refer to bronze sculptors, sthapatis (the same term is also used for other Kammalar artisans, including architects and masons), as hereditary temple servants, suggesting that at least some workshops were attached to individual temples. It is likely that, as for other crafts, considerable variation existed in the social and economic status of bronze casters, depending on their location, the quality of their work, their

205

206


hereditary positions, and relations to temple and royal patrons. However little in the way of specifics can be gleaned from either material evidence or Vijayanagara-period inscriptional sources.

Stone working: masons, sculptors, and engravers A number of factors contributed to marked increases in stone working over the course of the Vijayanagara period. New urban centers, temple towns, and settlements were founded and existing sites grew in size and architectural complexity. Settlements and temples were enclosed within massive masonry walls, and elite architecture and temples of all sizes involved stone construction in whole or in part. Often such structures were elaborately carved, requiring the skills of both masons and sculptors. Agricultural expansion into forested regions and the semiarid uplands around the capital involved the construction of numerous canals, wells, and irrigation reservoirs, while military activities led to the construction of numerous forts throughout the empire, as well as defensive features such as watchtowers and fortification walls. In addition, engravers were involved in the production of the many thousands of inscriptions that adorn the walls of Vijayanagara temples and, less often, occur on boulders and slabs throughout the empire. Numerous sculptors were required to decorate buildings and to carve freestanding images of deities, royal figures, secular scenes, and of individuals such as elite donors and male heroes who died in war, or heroic women who committed sati. Masons and sculptors belonged to the same community as smiths and carpenters: the fivefold Kammalar community discussed previously. As with smiths and weavers, considerable occupational differentiation appears to have existed among workers in stone before and during the Vijayanagara period. These include diverse categories of skilled artisans, as well as the various skilled and less (un?) skilled laborers engaged in large-scale construction activities. We have some information on the hierarchical relations among categories of artisans involved in architectural construction. A number of ancient Sanskrit and Tamil texts provide information on historical conventions for the form and production of specific architectural and sculptural forms. These texts date to the first millennium AD and include the Sanksrit Manasura silpasastra (The Science of Architecture), Mayamata silpasastra, Amsumadbheda, various Puranas, including the Matsya-purana, and the Skanda-purana, as well as Buddhist and other sources (see Acharya 1980 [1934]). Such texts would certainly have been


familiar to master architects and sculptors, if not to the majority of the artisans who performed much of the physical labor of construction. The texts contain prescriptions concerning both the forms and production techniques for temples and sculpture, and provide some information on the specialists involved in their production. In the Manasura the architect or masterbuilder (sthapati) is described as the offspring of Shiva as Visvakarman, the “great architect of the universe” (Acharya 1980 [1934]: 6). The architect should be knowledgeable in the Vedas (the ancient Sanskrit sacred texts) and in all architectural texts and “the science of architecture.” His role is that of director, responsible for supervising other personnel involved in construction. These include the sutragrahi, expert in architectural drawing, the vartaki, which Acharya classifies as an expert in painting, and the takshaka, or master carpenter (Acharya 1980 [1934]: 6–7). From a period more proximate to Vijayanagara, a twelfth-century inscription from Tiruvar temple (Tamil Nadu) refers to several categories of skilled artisans involved in temple construction (Ramaswamy 1993: 29–30). These included individuals who marked and measured stones before carving (sutragrahi), stone dressers or joiners (vartaki; a rather different definition of this task than provided by Acharya), and sculptors (shilpa or kal-techchan). Temple repairers (tirrupani cheyvar), engravers, and “craft overseers” (kankani) are also mentioned in Chola inscriptions (Ramaswamy 1993). From the Vijayanagara period, we find references to masons, temple repairers, and carpenters in numerous temple inscriptions. In all cases, these artisans are described as temple functionaries, and designated as receiving payments from temple donations. An additional category, engraver, is more ambiguous, and refers to individuals who carved lithic or copper plate inscriptions. I will discuss these separately below. Many other inscriptions refer to generic “artisans” (sippiyar, Tirruppaniseyvan), who may also have been involved with temple construction. The evidence thus indicates a welldeveloped division of labor among skilled stone workers working in temple contexts, and presumably in other elite contexts as well. In many cases, these artisans appear to have been members of a permanent temple staff, i.e., attached specialists, who received payment in food and cash for their labor (see chapter 7). Unlike the Kaikkolar weavers discussed earlier, the masons and carpenters associated with Tirupati, Tirumalai, and numerous other temples are not reported as having assumed additional roles in temple administration or temple activities (such as decorating mandapas or carrying sacred images). Nor are they listed as receiving ritual privileges or making temple donations.

207

208


Instead, these artisans were compensated through others’ donations for the performance of their traditional occupations. Although their labor was clearly in demand during the Vijayanagara period, stone workers do not seem to have experienced significant changes in their social or ideological status (though individual master artisans may well have). Although a range of diverse occupations associated with stone working is well documented from a variety of sources, reference to individual stone workers of the Vijayanagara period is comparatively rare. However, in a number of cases, names of builders, architects, sculptors, and engravers were recorded (see also pp. 232–234). There are several such inscriptions at the site of Vijayanagara. An inscription near a diversion weir (anicut) along the Tungabhadra River near the Virupaksha temple records that the construction of the weir was “caused by Chintayaka Devanna and was constructed by Bomoja” (summarized in Patil and Patil 1995: 16, inscription 56). An inscription on a split boulder in the Turtha canal at the capital, notes that Hegade Hitalabagila Basavana-anna “caused the construction (of the canal) by cutting the boulder . . . by the order of the king” (Patil and Patil 1995: 78, inscription 268). It is not clear precisely what Bomoja’s or Hegade Hitalabagila Basavana-anna’s roles were in the construction reported on – and it is likely they were supervisory, rather than providing the actual physical labor, which no doubt required many individuals. A third inscription, on a boulder near a small mandapa (columned hall) near Matanga Hill in the Vijayanagara urban core, records that the structure was built by the mason Basava (Patil and Patil 1995: 70, inscription 239), in this case listing both his name and occupation. More often, inscriptions refer to the individuals or groups whose investments sponsored particular constructions – whether temples, canals, columns, etc. – and not to the artisans who designed them or laborers who built them. In contrast to the elaborate division of labor described in the texts, a postVijayanagara source suggests that some stoneworkers performed multiple tasks. In a description of a stone quarry near Srirangapattana in the early 1800s, Buchanan (1988 [1807], vol. I: 133) reported that: “the same persons cut the stones out of the quarry and afterwards work them up into the various fantastical shapes that are given to them in Hindu buildings.” However, he went on to note that “the drudgery is performed by common labourers” (Buchanan 1807, vol. II: 133), suggesting at the very least a division of labor between skilled masons and unskilled workers. Despite the static nature of the prescriptive texts, Vijayanagara architecture was far from unchanging. Michell (1994, 1995) and Dallapiccola


(1998) have documented the diverse influences on and temporal changes in Vijayanagara architecture and sculpture from the fourteenth through seventeenth centuries. In both we see dramatic changes in style and in scale. Fourteenth-century temple architecture at the imperial capital (we know less of administrative architecture for this period, though the pattern is likely similar) drew on local styles of the northern Deccan region, and constructions were relatively small. By the late fifteenth and particularly, the early sixteenth centuries, a distinctive imperial style had been created, evident in both temple and administrative architecture. The scale of construction increased dramatically, affecting both the size of individual structures or complexes and the rate of their production at the imperial capital and throughout the empire. This period is also marked by an increased standardization of architectural and sculptural forms across broad imperial territories (Dallapiccola 1998: 142), as a “signature” Vijayanagara style emerged. In temple architecture, stylistic influence from the southern Tamil region became prominent and a widespread architectural unity emerged, evident in characteristic towered temple gateways (gopura) and elaborately carved columns. In administrative architecture, Vijayanagara architects drew on an even broader region in their creation of an imperial style, most particularly the northern Bahmani and its successor sultanates. This was part of a broader process of the creation of pan-South Indian expressions of political authority and kingship that linked Deccani polities and Vijayanagara in many symbolic and political domains (see Michell 1992a, 1995; Wagoner 1996; Sinopoli 2000, for discussions of this process). The diverse sources of stylistic influence and the increasing stylistic standardization that resulted from merging these influences into a distinctive Vijayanagara form clearly provide evidence for the movement of ideas over broad areas as well as the likely movement of skilled architects and artisans within and across political boundaries. In administrative architecture too, new forms were created that communicated Vijayanagara’s claims to domination in a distinctive and highly visible style. Michell (1992a, 1995) has written extensively about the emergence of Vijayanagara courtly architecture, arguing that it drew from both southern, “Hindu,” architectural traditions, and northern, “Islamic,” traditions. The latter extended to construction styles including the extensive use of mortar instead of dry-stone masonry and changes in the size of building blocks, as well as form, and ornamentation, particularly elaborate stucco and plaster work. Vijayanagara courtly architecture has close parallels in structures at the various Bahmani capitals in the prominence of domes,

209

210


6.11 Administrative architecture in the Vijayanagara urban core: royal elephant stables

arches, and other features. The emergence of this widespread administrative style has been interpreted by Michell and others as providing evidence of emergent, pan-regional concepts of kingship, authority, and royal presentation that linked these often antagonistic polities. The Vijayanagara courtly style evolved relatively early in the empire’s history, and Michell (personal communications) has suggested that some of the courtly structures at the Vijayanagara capital, e.g., the elephant stables (Figure 6.11) and a towered reception hall, popularly referred to as the Lotus Mahal, date to the early fifteenth century. The sheer scale of many Vijayanagara monuments and other masonry features demonstrates that large numbers of laborers were involved in their construction. This is supported in written sources (discussed in chapter 4) that indicate that enormous work groups could be mobilized for projects such as the construction of large irrigation reservoirs. Paes’ description of the construction of a reservoir in the Vijayanagara metropolitan region (see chapter 4, this volume; Sewell 1900) provides estimates of some 15,000 to 20,000 laborers engaged in its construction. While the accuracy of this estimate is certainly questionable, its scale nonetheless provides a sense that very large numbers of workers were mobilized for particular tasks. Similarly (and perhaps more realistically), the Porumilla reservoir inscription records that 1,000 workers labored for two years in the construction of the approximately


six-kilometer long embankment (Parabrahama Sastry 1978:10; Ramaswamy 1993: 30). As discussed in chapter 4, it is less clear how such work groups were coordinated, provisioned, or recruited. Paes noted that the labor in the reservoir construction that he observed was apportioned out among the king’s “captains,” suggesting multiple groups of laborers under multiple supervisors. Such a pattern may also be indicated in the architectural details of the city walls of Vijayanagara. Robert Brubaker (in prep.), the Vijayanagara Research Project, and the Vijayanagara Metropolitan Survey have recorded details of fortification wall construction in the urban core and metropolitan region. It is not uncommon to find adjacent wall segments with subtle differences in construction styles, including block size, block alignment and coursing style, and the presence of absence of chinking. In addition, bastions on fortification walls typically abut against the walls (rather than join them) and appear to have been constructed as separate segments. The precise sequence of wall construction is not known. However in order to serve their intended purpose, fortification walls and, presumably the bastions, must have been designed and built as coherent units and relatively rapidly (see Brubaker in prep.). While by no means conclusive, the wall construction evidence suggests that many different work groups may have been responsible for their construction. Such groups, even when building adjacent segments, may have adhered to slightly different stylistic conventions and building techniques. Still other groups of workers may have been responsible for the construction of the bastions. This interpretation of multiple work groups corresponds with Paes’ description of the multiple “captains” who supervised groups of laborers in the reservoir construction. Many Vijayanagara fortification walls were constructed of exquisitely crafted dry-stone masonry (Figure 6.12). At their best, they are constructed of vertical (or slightly sloping) faces of tightly fitted dressed stone blocks. The blocks, which range from about forty centimeters to more than two meters on a side, have rectilinear faces and taper in the rear (Figure 6.13), so that only their front faces need to be fitted. Fortification walls are either double-faced with rubble fill or, in many contexts, are faced only on the exterior and supported on the interior by an earthen embankment. The designers and builders of Vijayanagara fortifications also made skilled use of the natural features of the landscape – in their placement and in the incorporation of natural boulders or outcrop ridges into defensive systems. The construction of fortifications required artisans with a variety of skills. Architects or others expert in military architecture and military strategies would have been employed in design and placement. Skilled masons were

211

212


6.12 Dry-stone masonry in the Vijayanagara urban core: detail of wall construction

6.13 Section of fortification wall in the Vijayanagara urban core: trapezoidal blocks


likely responsible for quarrying and fitting the stone blocks. Less skilled laborers (perhaps soldiers during lulls in warfare?) could have been employed in transporting and lifting quarried blocks and in mining and placing earth and rubble fill or embankments.

Stone quarrying and construction The large granitic outcrops of the Deccan plateau provided abundant raw material for Vijayanagara masons and sculptors at the capital and in the northern area of the empire. There is extensive archaeological evidence for quarrying activities across the Vijayanagara metropolitan region. Not surprisingly, quarrying typically occurred near to building sites and quarried blocks rarely appear to have been transported long distances. In general, stoneworkers used the most proximate available raw materials. Thus, in areas of the metropolitan region where granite outcrops do not occur, locally available stones or earth were used for building fortifications. This is most notable in the outer areas of the Vijayanagara metropolitan region, where dykes of basalt (often in the form of small rounded cobbles and boulders) dominate in many outcrop hills. Here, fortification walls were typically built of the small rounded basalt or “gabbro” stones, and faced masonry block construction is rare (Figure 6.14). The walls, while functionally identical to their masonry counterparts, present a very different appearance. It is important to note that these walls tend to occur on the outskirts of the metropolitan region in areas that would have seen relatively little traffic, and gates on these walls, where traffic would have been heavier, are typically of distinctive Vijayanagara masonry. In upland sections, the “communicative” aspects of fortifications – as highly visible monumental expressions of Vijayanagara’s power and dominance – would have been less important than in the city core or along major thoroughfares. This suggests that both symbolic and economic concerns influenced the construction of fortifications. Evidence for stone quarrying is abundant throughout the metropolitan region. Quarry marks were recorded in fifty-six sites in the intensive survey region, and in many locations that were not assigned specific site designations. These marks take the form of small square depressions, three to five centimeters across and spaced between five and twenty centimeters apart, and were pecked into outcrops to define fracture lines along which outcrops were cut (Figure 6.15). Buchanan (1988 [1807], vol. I: 133–134) described similar techniques of stone quarrying near Srirangapattana in the early 1800s:

213

214


6.14 Fortification wall (VMS-339), constructed of unmodified boulders


6.15 Quarry marks incised on large outcropping boulder in Vijayanagara metropolitan region

The granite may be cut in wedges in any direction, and to any length; but there is always one direction in which it is found to split easiest; a number of small square holes, about an inch and a half in diameter and four inches deep, are cut in the line by which the stone is meant to be split. The work is formed by a small steel punch of this shape, which is driven by a heavy iron mallet. When the rock or stone is very long, or deep, these holes must be almost contiguous; but when the surface to be split is small, they may be at considerable distances. Blunt wedges of steel are then put in the holes, and each is struck in its turn, until the stone splits, which it does in a straight line to the very bottom of the mass or stratum. The surface is cut smooth with steel chisels, and, except, in the very finest works, receives no higher polish.

While, as discussed earlier, steel was produced during the Vijayanagara period, we do not know if steel tools were readily available to quarriers. Filliozat and Filliozat (1988: 3–4) have suggested that Vijayanagara stone masons may have instead used wood to help split stones. This would have entailed placing wooden plugs into the quarrying depressions and soaking them with water so that the plugs expanded, fracturing the stones. As noted, evidence for quarrying is common at a range of site types. A small number of sites identified in the Vijayanagara metropolitan region were identified as being exclusively or primarily quarrying sites. Site VMS341, located to the southeast of the urban core (Figure 6.16), is a small site of

215

216


6.16 Quarrying site in the Vijayanagara metropolitan region (VMS-341)

c. 160-square meters located on a sheetrock outcrop. Quarrying at this site appears to have ended abruptly, and the site consists of a several boulders that had been split into large rectangular blocks, but not yet trimmed into blocks for use in construction. A surface scatter of earthenware ceramics is associated with this site, suggesting a short-term encampment by stone workers. No Vijayanagara sites with masonry construction are found near this site. Whatever building project the quarrying was intended for, it was not completed. No temporally diagnostic evidence was recovered to identify when quarrying was abruptly halted, and while it is tempting to date it to the AD 1565 abandonment of the capital, this cannot be confirmed. A second quarrying site (VMS-423) consists of a zone of intensive quarrying and associated ceramic scatter in a saddle on a high outcrop. Notable at this site is a set of vertical tic marks inscribed into an outcropping boulder near the quarrying locale. This appears to be the remains of some kind of recording or counting system, perhaps for keeping records of rates of production or of laborers. Both of these sites are relatively small in scale and were likely used for brief periods by small mobile work groups. Another site in the metropolitan region contains valuable evidence for construction processes. This is a small, uncompleted shrine, located along the Turtha Canal to the east of the urban core (VMS-79). This four by three column structure was in the process of being built on a sheetrock


6.17 Uncompleted structure (VMS-79): blocked out column and trimmed blocks

surface along the face of large outcropping boulder when a portion of the overhanging outcrop collapsed upon on it, halting construction. Remains at the site consist of column footings that were pecked into the sheetrock to define the structure’s plan, a completed column crushed beneath the fallen overhang, and several partially completed columns (Figure 6.17). Also present were significant quantities of granite flakes and debris indicating that the columns were being shaped in situ. A small image of Hanuman was carved on a nearby boulder, and a two-character inscription was inscribed was on a boulder to the east of the structure. This site presents a rare glimpse of construction activities in progress, by a work group engaged in all stages of the construction process. While the dominant image is thus of groups of workers moving from site to site, there is some indirect evidence for the mass production of certain architectural elements. Columns in numerous sixteenth-century temples throughout the urban core and metropolitan region (and beyond) appear to have been pre-manufactured and subsequently modified to fit into the structure into which they were placed. This is not the case in early Vijayanagara temples, but becomes widespread in sixteenth-century structures, when temple size was increasing, and large numbers of temples were being rapidly constructed (Dallapiccola 1998). In the most common column type of this period, columns were divided into three square or rectangular panels

217

218


separated by octagonal insets. Modifications to columns either involved placing columns that proved to be too short for their intended structure on low stone bases (often just coarsely shaped granite blocks of the appropriate height), or topping them with two capitals instead of the usual one. For columns that proved too tall for the structure, parts of the upper panel were lopped off. For sculpted columns, this meant images carved on the uppermost column panels were partially cut off. No attempt was made to modify or resculpt the affected panel, resulting in representations of headless or partial deities or a portion of a medallion or animal sculpture (Figure 6.18). It is important here to note that most or all of these columns would have been covered in plaster and painted and some corrections (or masking) of truncated images could have occurred at that stage. These trimmed or modified columns occur in numerous structures throughout the urban core and metropolitan region, from large royally sponsored temples to much smaller shrines and temples. The widespread pattern suggests that column types were being manufactured according to a general standard rather than for particular buildings or locations within buildings. It is still likely that manufacture still took place relatively near the structures in order to minimize the costs of transporting heavy columns. But the impression is of rapid and relatively large-scale production by work groups of specialist column producers – including quarriers, column shapers, and sculptors (tasks that may have been performed by the same individuals), rather than of highly planned customized construction of individual buildings by a coordinated team. A small carving on the walls of Raichur fort to the north of Vijayanagara provides some evidence for the logistics of movement of large stone blocks (Figure 6.19; Sugana Sarma 1998: 87; Basham 1954: 219). This simple schematic carving (graffiti?) depicts a large sculpted column being transported on a wheeled cart, pulled by oxen and aided by several men. The precise date of this inscription is unknown, and the fort spans the Vijayanagara period to the present, but the carving style suggests that it could date to the Vijayanagara period. This unique image presents a rare glimpse of non-elite physical labor in historical South India.

Sculpture Sculpted stone images of the Vijayanagara period take a range of forms. They include images of deities carved on slabs and boulders; sculpted columns, basements, wall panels, and images carved on reservoir sluice gates or other constructions; and sculpted stelae or columns that marked


6.18 Truncated columns in the Vitthala temple complex, Vijayanagara urban core

territorial boundaries or honored heroes who died in war and women who committed sati. Vijayanagara-period stone sculptures range from elegant finely executed works of extremely high artistic quality, to much simpler and less well-carved images. Some of this variation is a product of the different stylistic traditions employed for specific sculptural forms and temporal changes in sculpting styles (see Dallapiccola 1998). However, it also results from the widely varied skills of individual carvers or carving “workshops.”

219

220


6.19 Carving on the walls of Raichur Fort depicting transport of large quarried column (after Sugana Sarma 1998: 87)

Vijayanagara sculpture has not been a major focus of traditional South Asian art historical literature, being considered less noteworthy or artistic than the sculpture of earlier periods, though Michell (1995, chapter 6), Verghese (1995, 2000), and Dallapiccola (1998; Dallapiccola and Verghese 1998) have provided important recent discussions of Vijayanagara sculpture and iconography. Here my concern is with the production and producers of stone images, rather than with a detailed discussion of forms or stylistic content. Along with large non-portable or semi-portable images, countless smaller portable stone images were also produced during the Vijayanagara period. Many of these were related to religious worship. Small stone images may also have been produced for sale at pilgrimage sites. The mid-sixteenth-century text, Mohanatarangini (Chariot of Krishna), written by the Vijayanagara provincial ruler Kanakadasa, refers to sculptures for sale in the markets of the imperial capital (Kotraiah n.d.d: Book III, lines 18–20). These small images were made in a variety of raw materials, including chlorite, crystal, and a wide variety of semi-precious and precious stones. While some of these miniature images have been recovered in excavations or are found in museum and private collections, we know little about their production. Given the range of stone sculptural forms produced and their diversity of functions, it is probable that the organization of such production was organized in a variety of ways and at a variety of scales. Buchanan’s description of stone working near Srirangapattana, quoted previously (p. 208) suggests that in some contexts, stone carving may have been carried out by members of the same work groups engaged in stone quarrying. Such may certainly be case for the “mass-produced” sculpted columns discussed earlier. Sculptors sometimes signed their work. An undated inscription on a boulder in the Tungabhadra River of the fifteenth or sixteenth century notes that the nearby images of Ganesa and Hanuman were carved by Lakhanna,


the son of Devanna of Varanasi (Patil and Patil 1995: 46–47, inscription 163). If this is the same Varanasi that lies on the Ganges River, then this artisan came from a very long distance indeed. In general, inscriptions naming sculptors are relatively rare. As discussed below, signatures are associated with several memorial stone images of the period, primarily in western India. Individual names or letters are also sometimes inscribed on sculptures (see Patil and Patil 1995: 126, inscriptions 433 and 434), but it is difficult to determine if these refer to the sculptors, patrons, or purchasers of the image, or to the image itself. In the following, I discuss various categories of images separately.

Sacred images Hindu worship involves the veneration of sacred images believed to be imbued with the presence of the divine. Dozens of images – of stone, bronze, and wood – can be found within large temples. These include the central non-portable image of the temple’s main deity, as well as numerous smaller portable images of the same god. These portable images are used in processions, chariot festivals, swing ceremonies, and other rituals where the god exits the sanctuary to come into the presence of his or her devotees. Representations of subsidiary deities are also common and include representations of counterparts or spouses of the primary deity, guardian figures, and sacred vehicles (such as the bull Nandi in Shaivite shrines, or the eagle Garuda in Vaishnava ones), or other gods related to the temple’s central deity (e.g., such as the presence of Hanuman images in temples dedicated to Rama). Temples may also contain images of diverse manifestations of the temple’s central deity (various forms of Shiva or Vishnu). Although temples are typically dedicated to one deity, many pre-1529 AD Vijayanagara-period temples contained both Shaivite and Vaishnavite images, particularly in column sculptures. This pattern changed during the reign of Achyutadevaraya and his successors when Vaishnava temples and images were increasingly promoted (Verghese 1995), and Shaiva images and temples disappear from imperial constructions. Sculptors associated with temples may have also produced portable images for devotees and pilgrims. In many contexts, master sculptors or sculpture workshops may have been resident on temple grounds, as attached specialists. It is also likely that independent sculpture workshops were present in temple towns and near temple precincts. Several sacred texts refer to the sculpting of sacred images for temples and shrines. These suggest that the entire process of image creation, from

221

222


quarrying to installation, had a powerful sacred dimension, as the presence of the deity became increasingly manifest in the sculpture throughout the process of its formation (Davis 1997: 33–35). Davis (1997) documents the interaction of priests and sculptures in the process of “establishment” (pratistha), whereby the deity comes to inhabit the image as a living presence. This involves a series of ritual acts that begin with the selection of the raw material, continue throughout the carving sequence, and culminate in the “awakening” of the image through the ritual opening of its eyes. Evidence from sacred texts thus suggests that the carving of images destined for particular temples or sanctuaries involved a close relation between priestly patrons and artisans. While the production of sacred images for major temples no doubt involved complex relations between patrons and producers, many other Vijayanagara-period sacred images occur in non-temple contexts, and we have little knowledge of their production. Some of these images are related to fortifications and other “state” constructions. These include the many images of protector deities, often fierce manifestations of Shiva, such as Bhairava or Virabhadra, that are common at boundaries, gateways, and fortifications (Michell 1995: 166). Images of Hanuman (Figure 6.20) are also common at boundaries throughout the metropolitan region, making explicit the site’s sacred associations with Hanuman’s homeland Kishkinda. Hanuman and Ganesha, both well loved and generally benevolent deities, as well as other gods, goddesses, and sacred images (such as naga stones, or snake images that may be associated with fertility, see Figure 6.9) are common in settlements and along roads. And isolated images often occur in agricultural fields and on hilltops, sometimes, but not always, in temples or constructed shrines. A broad range of gods and goddesses are represented in iconic and aniconic forms, from formal Sanskritized forms of Vishnu and Shiva to local deities, worshipped by the region’s diverse inhabitants. Stone images were carved on outcrop faces and bedrock boulders, indicating that landowners, state officials, village organizations, or other sponsors commissioned sculptors to come to sites. Most often, the images were freestanding, and carved onto granite slabs. Many of these comparatively “portable” images (though some were quite large and weigh a ton or more) have been relocated since Vijayanagara times. In the metropolitan region, they are often found in modern structures in settlements or along roadsides; many Vijayanagara images are also found in archaeological museums or private collections. Their original context of use is thus lost. Nonetheless, it is apparent that the


6.20 Finely sculpted Hanuman image on freestanding slab (VMS-202)

complex distributions of images and sacred spots made the Vijayanagara metropolitan region a heavily marked sacred landscape (Sinopoli 1993). The images found throughout the metropolitan region vary widely in terms of the skill or energy invested by their carvers. Some are highly refined, of exquisite workmanship with elaborate ornamentation, and were clearly carved by skilled artisans trained in classical conventions (Figures 6.20, 6.21). Others are much coarser, poorly proportioned, with coarse features or “folk” themes (Figure 6.22).

223

224


6.21 Finely sculpted goddess image on outcrop

6.22 Crudely carved depiction of Rama and Sita images on boulder in Vijayanagara metropolitan region


6.23 VMS-538: Unfinished Hanuman image

A few sites documented by the VMS provide evidence for the process of sculpting. These three sites contain unfinished images – one of Ganesha (the elephant-headed son of Shiva) and two of Hanuman. The Ganesha image (VMS-94) was being carved on a small boulder (2 × 3 m) in a field located on the edge of the drainage basement of a reservoir embankment to the east of the urban core. The small image (.75 m tall) of the deity was surrounded by an arch above which was sculpted a crescent moon and sun. While the Ganesha was fully carved, the arch and celestial images had not yet been completed; only their outlines were pecked into the boulder. Beneath the image was an inscription within a square frame. Unfortunately, the stone was heavily exfoliated and the inscription could not be transcribed. The first uncompleted Hanuman image (VMS-567) also occurs on a large boulder, in this case located in a small Vijayanagara-period settlement with several rubble wall structures and a circular bastion. The outlines of a Hanuman image have been pecked on to the boulder (c. 2 m × 1.5 m), and carving had begun; we do not know why it was never completed. A second uncompleted Hanuman image (Site VMS-538; Figure 6.23) was halted when a large crack developed in the slab on which it was being carved. This unfinished carving sits in an area of outcrop and quarrying. It is located along the route of a Vijayanagara-period road and less than 100 meters from an ancient settlement site. It is likely that the image was intended to be

225

226


set up very near to the location of sculpting. Several stages of production are evident here, including the quarrying and trimming of the 2 by 1.5 m slab. At the time production ceased, the slab had been only roughly shaped and would have required further trimming to stand upright. After quarrying and rough trimming, the outline of the image was roughed out on the slab, and then finer shaping took place. Only Hanuman’s right foot remains in outline form. Before it could be completed a crack developed in the slab near the knees of the image, and the carving was abandoned. All three of these unfinished images appear to be located near the site of their intended disposition. Their production likely involved either resident or mobile artisans, who were recruited to complete individual sculptures. Given the enormous numbers of sacred images across the Vijayanagara, there must have been many such sculptors, and their work was likely in considerable demand.

Other images Portraiture Other categories of sculpted images of the Vijayanagara period paid tribute to specific individuals, whether royal figures, elite donors, heroes who died in battle, or women who committed sati. I first discuss royal and elite images, before turning to the latter two categories of images. Royal portraits occur on temple walls or columns, including in the Tiruvengalanatha and Krishna temples in the capital and in other temples throughout the empire (Dallappiccola 1998). Royal figures are represented throughout the Vijayanagara sequence; however, the nature of their representation changed over time. Fourteenth- and early fifteenth-century images, on the royal throne platform or the enclosure of the Ramachandra temple in the capital’s royal center, portrayed paradigmatic seated royal figures receiving merchants or tributaries in processional scenes (Figure 6.24). These do not seem to be attempts to portray individual kings, but are instead generic representations of royal authority. A small mid-fifteenth-century Shaiva temple in the Vijayanagara urban core contains two portraits of named individuals, the king Mallikarjuna (AD 1446–1465) and his attendant, Sirangu, who sponsored the temple’s construction (Verghese 2000: 48); these are also fairly stylized images, identifiable through their associated inscriptions. By the sixteenth century, these kinds of portraits of specific rulers became common (Dallapiccola 1998). They occur in the capital on the Krishna temple (depicting Krishnadevaraya) and the Tiruvengalanatha temple (perhaps depicting Hiriya Tirumala, brother-in-law of Achyutadevaraya; Dallapiccola


6.24 Seated royal figure sculpted on the face of the royal throne platform, Vijayanagara urban core

1998: 151). In these contexts, rulers are depicted in worshipful poses, standing before an image or with their hands folded in front of them (in the anjali mudra). Other elites, individuals who sponsored the construction of temples, reservoirs, or other facilities, are depicted in a similar fashion in “donor portraits” (see Figure 6.7). Inscriptions are not typically associated with these images, so we seldom know the specific individuals depicted, but their garments reveal them to be of high or royal status. Royal portraiture expanded even further in the seventeenth century (Michell 1995: 180–184), during the period when the power and centralized authority of the Vijayanagara empire was waning and numerous claimants to royal authority appeared. Like the expansion of patronage of poetry discussed earlier, the spread of royal portraiture in sculptural idiom is probably best seen in the context of this new political dynamic. Most striking are the labeled portraits of ten rulers that were carved on the Pudu Mandapa at Madurai, sponsored by the powerful Nayaka ruler Tirumala (r. 1623–1659). The rulers are labeled and placed in chronological order, starting with Vishvanatha (r. 1529–1564), a former officer of Krishnadevaraya, and ending with Tirumala (Michell 1995: 17–19); together they trace the ruling lineage of Tirumala. Through these images, a powerful statement was made of historical depth and political legitimacy in a period

227

228


when both were heavily contested (see Sinopoli 2001 for a discussion of a similar practice in the much earlier Deccani Satavahana empire). In fact, in general, the seventeenth-century “Nayaka period” is marked by an increased size, elaboration, and ornamentation in temple architecture, administrative architecture, and sculpture, a potent reminder that material statements of power need not correspond with political realities.

Memorial stones Memorial stones – sculpted tributes to deceased warriors, Jain ascetics, or women who committed sati or ritual self-immolation – have a long history in South India, minimally spanning from the sixth through the eighteenth centuries. These sculpted stelae contain stylized depictions of individuals who met noble deaths. They were typically commissioned by relatives or followers and placed in the precincts of Hindu and Jain shrines and along roadsides and transport routes throughout South India. They provided a public and visible tribute to the deceased. Tens of thousands of memorial stones are found throughout South India; the vast majority date to preVijayanagara periods. Most memorial stones are uninscribed and have been dated based on stylistic evidence. Some stones have inscriptions that provide information about the deceased. And in a number of cases, they include information on their sculptors and inscribers. S. Settar and Gunther Sontheimer and their colleagues (Settar and Sontheimer 1982) have carried out the most comprehensive study of memorial stones conducted to date. They have documented 609 inscribed memorial stones from Karnataka, though their sample largely comes from the western regions of the state and does not include the area around Vijayanagara. Of the 609 inscribed stones, c. 122 date to the Vijayanagara period (20 percent), c. 124 date to the Chalukya period, and 363 to the Hoysala period (Settar and Ganihar 1982: 321). Approximately 137 sculptors are named; 77 (56 percent) belong to the Vijayanagara period. The reasons underlying the disproportionately high number of signed memorial stones of the Vijayanagara period are unclear; though this does seem to have been a trend in western Karnataka in particular, perhaps relating to the particular repute of individual carvers or workshops. Hero stones (viragal) or sculpted panels dedicated to deceased warriors comprise the most numerous type of memorial stone of the Vijayanagara period (Dallapiccola 1998). The most elaborate of the Vijayanagara hero stones contain three or four tiers of images. The lowermost tier depicts the hero’s death in battle; in the next tier, he is shown being transported to heaven, and in the upper tier, he is shown worshipping a seated figure


6.25 Memorial stone from Bhatkal, Western India

or linga (Figure 6.25). The details of the battle scenes suggest that these elaborate images were carved to recount the acts of a specific individual. These acts are, as noted above, sometimes described in lengthy inscriptions (Settar and Sunkad 1982: 344). Other hero stones are much simpler and more generic in content. They consist of a single panel depicting a sword-wielding warrior on horseback (Figure 6.26). Such stones carried the essential meaning of the more elaborate forms – a tribute to a heroic warrior – yet lacked detailed information concerning a specific individual and his particular death. We do not have sufficient data to consider temporal changes in the form of representation on hero stones. In the absence of evidence to the contrary, it may be suggested that these simpler hero images may have been commissioned by those unable to afford the more elaborate stones or may have been produced in larger numbers, rather than being manufactured on demand to honor a particular individual. If, as seems likely, these simpler stones increase in frequency over time, their appearance may relate to an expansion of claims to warrior or heroic status on the part of those with fewer economic resources than the elite and affluent families who commissioned the more elaborate stones. While dying in battle was an honored death for a male, the female equivalent was the act of sati or self-immolation on the funeral pyre of her husband. Sati stones depict such heroic women. These stones typically

229

230


6.26 Simple memorial stone of warrior on horseback in the Vijayanagara metropolitan region

depict the woman standing with her right hand elevated above her head (Figure 6.27). Her husband, if depicted, is shown smaller in scale with hands folded in acknowledgement of his wife’s sacrifice. The name of the woman is typically not recorded, nor do we learn about the circumstances of her death (whether voluntary or coerced). Like the simple hero stones, the sati stones identified in the Vijayanagara metropolitan region are simple and generic, and are not detailed portraits of individual women. The family of the sati likely commissioned and installed the stones in tribute to their relative and as a public claim to piety and devotion. Dutch traveler William Methwold described a less durable memorial that he witnessed to a woman who died at her husband’s funeral (in this case, by burial) in Golconda in the early 1600s: Yet after we came in, one of them [her friends] stroke upon the grave, laying his head close unto it and calling her by her name, and told she answered and expressed her content in the course she had taken. Over whom there was a erected a little thatcht [sic] cover, and her kindred not a little glorified in being allied to so resolute and loving a wife. (quoted in Moreland 1931: 29)

Sati stones are relatively rare at Vijayanagara; Verghese (2000: 135–136) identified only fifty such stones at Vijayanagara (in the urban core and the local museum), and the VMS recorded only one sati stone among the many


6.27 Memorial stone dedicated to a sati, Vijayanagara metropolitan region

hundreds of stone images documented. Verghese has suggested that, despite the attention paid to it by European visitors, sati was in fact uncommon during the Vijayanagara period. Other memorial stones of the Vijayanagara period pay tribute to Jains who committed one or another of two kinds of ritual suicide: sallekhana (starving oneself to death in fulfillment of a vow) or shagamana (ritual immolation by women, similar to sati). These stones are most numerous

231

232


in the western regions of Karnataka in the Konkan region, where they are found in considerable numbers in Jaina shrines and monasteries. While most Vijayanagara-period memorial stones were not engraved with the name of their carver, as noted earlier, a small number were inscribed. Settar and Sunkad (1982) have examined evidence for approximately twenty individual carvers of fourteenth- through sixteenth-century Karnataka. The majority of these come from western coastal districts (Shimoga and Karwar Districts) of the state. In the absence of more systematic research elsewhere, it is unclear whether such carvings were in fact more numerous in those regions or merely have been better documented. Information on the artist typically includes the sculptor’s (or inscriber’s) name, as well as the name of his father, and/or a maternal uncle, and/or his teacher. Sometimes the names of local rulers and the artist’s place of origins are also mentioned. A number of carvers are known from multiple inscriptions. Based on their studies of the locations of the inscriptions, Settar and Sunkad have argued that memorial stone sculptors did not typically travel far from their birthplace, a pattern also evident in earlier Hoysala and Chalukya periods. These were not highly mobile artisans, but instead appear to have been fairly tightly tethered to a single community or cluster of communities. That sculpting of memorial stones, and presumably other images, was a hereditary occupation is evident from several inscriptions that acknowledge a carver’s teachers and students. A late fourteenth- to early fifteenth-century carver named Madoja-Nagoja, known from three inscribed carvings honoring Jaina holy men, trained his son Banadoja as a sculptor (Settar and Sunkad 1982: 339–340). Another stoneworker and his kin labored as both sculptors and warriors. This individual, Timmayya, lived in the late fifteenth and early sixteenth century and is named in three inscribed hero stones (dating to AD 1494, 1522, and 1530). The stones were dedicated to three warriors of a local chiefly family, the Kaluvana of the Konkan coastal region of western Karnataka (Settar and Sunkad 1982: 341–342). Timmayya’s genealogy is presented in the inscriptions. He is described as having been born in the settlement of Kaikini, the son of a military leader Petala Timmayanayaka, and grandson of Kaluvana Timmanayaka of Govura Fort (the latter was presumably named after the Kaluvana family that he served). Both his father and grandfather were also described as stone carvers. The inscriptions note that Timmayya, too, was a warrior of Govura. Artisans sometimes worked in pairs to produce memorial stones. These include pairs of sculptors or collaborations between a sculptor and an engraver (Settar and Sunkad 1982: 344–345). From c. AD 1415–1425, the


sculptors Isaracari (son of Kesavacari; the appellation acari denotes their membership in the acharya or Kammalar caste, see p. 190) and Maniyacari (son of Ramacari) of western Karnataka (Haduvalli region) worked together on a number of memorial stones, though their precise division of labor is unknown. Sometime in the late 1420s, this partnership apparently dissolved, and later signed sculptures by these individuals were produced either singly or in collaboration with new partners. The inscriptions provide little information on how carvers of memorial stones were paid for their services or about their social status. Only one fourteenth-century inscription refers to payment for a memorial stone, noting a cost of 12 gadyana (varaha) for carving and consecration of the image (Settar and Sunkad 1982). While this seems a not insubstantial sum, without additional information on payment to sculptors of other kinds of images, it is difficult to evaluate this information (indeed, Settar and Sunkad imply that it is a very small payment). Unlike memorial stone carvers of the earlier Hoysala and Chalukya periods, the names of Vijayanagara-period carvers do not appear with honorific titles, suggesting that values attributed to this occupation may have declined over time, as indeed absolute numbers of memorial stones of all categories also decreased.

Inscriptions As discussed in chapter 5, Vijayanagara inscriptions occur in a range of contexts, most commonly on temples, but also on boulders and boundary markers, and as graffiti. The sheer numbers of inscriptions dating to the Vijayanagara period, particularly in temple contexts, indicate that numerous people were involved in their production. Engravers were common temple employees and are sometimes mentioned as recipients of temple donations. No doubt numerous engravers were also employed in courtly contexts, and engaged in the production of copper plate inscriptions. And we know nothing of the many scribes who wrote on palm leaf and other manuscripts. Engravers sometimes appended their names to the end of inscriptions, and are sometimes named as witnesses to transactions documented in the texts. In a number of cases, the engravers of inscriptions are distinguished from the texts’ composers, and both are mentioned by name. These composers included individuals described as temple accountants or poets. The latter seem to have especially been involved in the authoring of royal texts, both panegyrics and accounts of royal donations. As discussed earlier, the court poet Sabhapati authored several lithic and copper plate inscriptions

233

234


of Krishnadevaraya. A seventeenth-century copper plate from the temple in Chamarakanagara, Mysore District records a donation by a local ruler named Chikkadevaraya and notes that the text was composed by the court poet, Tirumalayarya, son of Alagasingaraya, author of “excellent speech pregnant with Vedic wisdom” (EC IV 1975: 593–4). For non-royal inscriptions, temple accountants seem to have been the most common authors (e.g., EC IV 1975: 619–620); in other cases, only the name of the composer is provided, with no additional information on primary occupation. An interesting signature appears on a copper plate from Mysore district that records a grant made by Saluva Narasimha (AD 1493). This inscription notes that the plate was “engraved by the smith Bhairavacharya, son of Tippayacharya, a good sculptor” (MAR 1924–111, EC III 1974: 811–815). While both father and son were members of the fivefold Kammalar caste, they appear to have engaged in somewhat different occupations, suggesting individuals pursued diverse activities within the range open to this broad hereditary community. An intriguing trend in inscriptions of the Vijayanagara period is the appearance of unreadable or garbled inscriptions. These inscriptions are interpreted as having been carved by individuals who were not fully literate. One such inscription was documented at a small temple complex in the Vijayanagara metropolitan region (VMS-111, Figure 6.28). This undated inscription has been stylistically assigned to the late fifteenth or early sixteenth century. It records the erection of a column (?) dedicated to the god Mailara, by an individual named Hona-aya (Patil and Patil 1995: 113, inscription 391). However, further details are obscure, as many of the characters in this four-line inscription are unreadable. This appears to result from errors made by the inscriber in transcribing a text (originally written on paper or palm leaf, or perhaps in charcoal or chalk on the stone itself) by a literate composer. This in turn suggests that the inscribers were not literate in the language being inscribed. According to Patil (personal communications), illegible inscriptions become increasingly common in the later Vijayanagara period. Their occurrence suggests that lithic inscriptions were becoming important in a broader range of contexts than previously, as larger numbers of diverse individuals sought to leave permanent records of their generosity and socioeconomic status. It is also likely the case that a wider community of individuals may have been producing inscriptions in contexts outside of the control of large temples and their professional scribes and carvers. Such unreadable inscriptions also suggest that the donors whose acts are recorded may have been themselves illiterate, or else lived in distant locales and did not see their inscriptions.


6.28 Illegible inscription at small temple complex in Vijayanagara metropolitan region

Lithic artifacts: pegs and mortars Other kinds of stone working identified in the Vijayanagara metropolitan region include the production of stone artifacts for domestic tools and building materials. Such objects include grinding stones and pestles and stone pegs used in construction activities. Numerous structures across the Vijayanagara region were built on top of granitic sheetrock or boulders. Their walls were supported by flaked basalt and granite pegs (approximately 12–15 cm high by 4–6 cm wide × 2–4 cm thick), which were placed into square holes pecked into the outcrop. Evidence for peg production, in the form of completed pegs and/or flaked basalt debris, has been found at four sites in the Vijayanagara metropolitan region (Table 6.3). Three are settlements with restricted areas of peg production, and include a small village (VMS-35–37) and large town site (VMS 2) in Block O, and the late Vijayanagara suburb associated with the Achyutadevaraya in Block S (VMS-140). The third area of basalt working was also associated with this late burst of urban growth to the southeast of the city core (VMS-172). Basalt working within settlements appears to have been restricted to limited neighborhoods or households, suggesting relatively small-scale workshops, perhaps primarily for local use.

235

236


Table 6.3 Peg production in the Vijayanagara metropolitan region Site

Dimensions

Description of Remains

VMS-2

390 × 100 m

Large settlement site (old Venkatapur?), with basalt pegs, spheres, and debitage indicating basalt working. There is no sheetrock in this area of the site, suggesting that pegs were being produced for use elsewhere.

VMS-35

200 × 170 m

Settlement site. Basalt working concentrated in southwest area of site, along modern road. Remains consist of basalt pegs, handstones, and a scatter of flaked basalt debitage.

VMS-140

195 × 135 m

Settlement site, part of extensive surface scatter located in sixteenth-century suburb to southeast of Vijayanagara urban core (Varadadevi-amana-pattana). Site includes rubble walls, displaced sculpture, and artifact scatter, including flaked basalt debitage.

VMS-172

87 × 35 m

Platforms along Penukonda road. Several flaked basalt fragments found in association with platform.

Granite and basalt flakes are also found in the Vijayanagara urban core itself, though these have not been systematically documented. Ground stone mortars and handstones, used in food preparation are common in settlement areas in the urban core and metropolitan region. These mortars take two forms: nonportable mortars carved into sheetrock (Figure 6.29), and portable block mortars. The grinding surfaces of these mortars were similar. A c. 25–30 cm square outline was first pecked into the top of the stone; in the center of this, there was a circular depression approximately 15 cm in diameter and about 15–20 cm deep. Mack identified a mortar production area in the Vijayanagara city core near the sixteenthcentury Vithala temple (Mack 2000: 187–189). This production area covered approximately sixty square meters and contained five block mortars in various stages of preparation.

Other flaked stone tools Despite the importance of metal tools in South India from at least the first millennium BC, evidence from the Vijayanagara Metropolitan Survey indicates that the use of flaked stone tools persisted well into historic times. Flaked stone tools, typically simple flakes with minimal retouch, were found at more than a dozen sites in the metropolitan region. Materials used in


6.29 Completed and uncompleted bedrock mortar (VMS-101)

flaked tool production include quartzite, granite, and glass (Lycett 1991). Sites containing such remains include surface scatters of artifacts, which are admittedly difficult to date, but also well-dated Vijayanagara-period settlement sites, such as VMS-2. No formal tools were recovered. Instead, the industry can be characterized as highly expedient, in which simple flake tools were quickly prepared to perform specific cutting tasks. Given the ready availability of stone raw materials through the region, such artifacts were likely rapidly discarded following their use, and were manufactured on an as-needed basis. It is doubtful that any craft specialists were involved in this kind of production activity. Instead, I expect that stone tools were produced as needed by those categories of individuals who did not have ready access to metal tools. This may have included agricultural laborers, pastoralists, or other low-status laborers engaged in a variety of tasks. Other producers of flaked or shattered stone artifacts may have included smelters, who added quartzite to the walls of their furnaces to increase their ability to withstand high temperatures, and woodworkers.

Woodworking Carpenters comprise the last of the five groups of the Kammalar community. As with the other artisans discussed, they varied in skill and in the range of

237

238


objects they produced – from domestic and agricultural tools, such as plough parts and bullock carts, to household furniture, doorways, and musical instruments. Other carpenters were important in the production of ritual objects – elaborately carved temple guardians and chariots, as well as smaller portable wooden images. Today, the latter carpenters, particularly those who carve temple chariots, are highly specialized and travel to temple centers that commission their skilled labor to produce new chariots or repair existing ones. In many of their activities, carpenters must have worked closely with other Visvakarma artisans, particularly the iron smiths who produced the metal rings for cart wheels, plow shares, and chariot axles, etc., as well as the metal tools used in woodworking. Few well-dated wooden objects of the Vijayanagara period are known, though there are a larger number of wooden images and other items attributed to the seventeenth- and eighteenth-century Nayaka periods. Carpenters are mentioned in temple inscriptions (e.g., TTDES IV, GT326, p. 127; TTDES IV, TT544, p. 110–111; TTDES IV, TT140, pp. 175– 176; TTDES IV, GT344, p. 250), where they are listed as recipients of proceeds derived from temple offerings in exchange for their service as carpenters (the precise tasks this labor entailed are not specified).

Potters and pots Use and social significance While ubiquitous in the archaeological record of the Vijayanagara capital, earthenware ceramics and the potters who produced them are largely invisible in Vijayanagara-period texts. Ceramic vessels were necessary and widely used by individuals of all social strata, but pots and their producers had low social and ideological status. There is no evidence that potters experienced significant changes in their social or economic position during the Vijayanagara period. Both elite and non-elite households used ceramic vessels in domestic activities involving food preparation and consumption, and in household rituals. Ceramic well linings, roof tiles, and drainpipes were also used. Institutions, such as temples, military barracks, and administrative offices, needed vessels for storage and preparation and for storing and serving water. Temples also required ritual vessels. Vessels were sometimes gifted by donors for use at roadside wells; as discussed below, this is one of the rare contexts in which ceramics appear in written sources (e.g., Patil and Patil 1995, inscription 423, p. 123). Ceramic fixtures and implements were


used in industrial facilities, serving as tuyeres in iron smelting furnaces and as crucibles for copper smelting. Despite their widespread use, no locally produced ceramic wares or forms can be considered elite wares or status goods (imported East and Southeast Asian porcelain and stonewares were recovered at Vijayanagara, but lie outside the scope of this discussion). The low status of earthenware can be attributed to cultural values associated with beliefs about purity and impurity. In Hindu South Asia, ceramics have long been believed to be particularly vulnerable to the absorption of impurities. In this respect, they are deemed inferior to vessels of precious metals and of copper or bronze. This ranking of materials has a long history that can be traced through Hindu sacred texts for two millennia. According to the Dharmashastra – a set of codes or laws for right behavior and religious practices believed to have been compiled in the first two centuries AD (Thapar 1966: 121) – earthenware vessels absorb the nature or quality of the cooked foods that they hold (and their method of preparation), as well of the persons who touch them. To avoid pollution, members of ritually high-status groups should not accept vessels containing cooked foods or water from individuals belonging to groups of lower social or ritual status. Once polluted, vessels can in most cases be purified, but here again this is most difficult for ceramics. The Dharmashastra informs us that simply washing them in water can purify polluted gold and silver vessels, while brass and copper vessels must first be scoured with ashes. In contrast, earthenware vessels must be immersed in fire and, in some cases, can not be cleansed at all and must be discarded (Kane 1973: vol. IV: 315–326). Because of these prohibitions, Hindus do not typically dine off of ceramic vessels, but instead use either metal vessels that can be cleansed, or banana leaves that are discarded after a single use. The translation of a song by the sixteenth-century composer Purandaradasa (AD 1485–1565) reveals the low economic value of earthenware ceramics during Vijayanagara times: Please dear sister fetch me the pitcher used for bringing water from the river. If this pitcher happens to shatter it costs just a penny so bring it hither . . . Please . . .

239

240


Please bring the pitcher Used to fetch the Rama nama water Fresh and delicious I want to spend time with the water fetchers the women who carry this water in pitchers . . . so please . . . Please bring the pitcher which is used to carry the water known as great Govinda Hari I’ll pour ablutions on lord Purandara sitting on the hill-top Lord Bindu Madhava . . . so please . . . (quoted in Jackson 1998: 144–145)

The negative symbolic associations of ceramic vessels and the absence of ceramic serving forms that functioned in public feasting, gifting, or display contribute to the challenges of interpreting archaeological ceramics of the Vijayanagara period. Nonetheless, the close relation between ceramic forms and food practices may provide an avenue for examining at least one dimension of their social meaning (and spatial structure) – caste and community identity (Sinopoli 1993b, 1999). This is expected to be the case if different groups selected different ceramic forms from within the range available from producers, and if these forms were associated with particular dietary practices, or other dimensions of identity.

Ceramic forms Ceramics of the Vijayanagara period have been studied in detail only at the imperial capital and its metropolitan region (see Sinopoli 1986, 1988, 1993b, 1999). Small numbers of ceramic vessels have been illustrated in archaeological reports from other regions of the empire. Although no systematic studies have been conducted on the latter, vessel forms appear broadly comparable to those identified at the Vijayanagara capital. It is nonetheless important to emphasize that detailed information on ceramic materials and production technologies is all derived from one small (albeit important) region


of the empire, and generalizations to broader imperial territories remain problematic. At first glance, Vijayanagara ceramics appear tremendously uniform. The ceramics are mostly (more than 90 percent) dark brown or black in color. They are undecorated or have simple impressed decorations. Surface treatment is minimal – most vessels were lightly smoothed or polished, but highly burnished wares are relatively scarce (

The Political Economy of Craft Production: Crafting Empire in South India, c. 1350-1650

Crafting the Nation in Colonial India

Craft Of Political Research

The Knowledge Economy in India

The Political Economy of Slavery: Studies in the Economy and Society of the Slave South

India: macroeconomics and political economy, 1964-1991

North and South in the World Political Economy

The Political Economy of Work (Routledge Frontiers of Political Economy)

Peasant History in South India

The Political Economy of Commerce: Southern India 1500-1650 (Cambridge South Asian Studies)

Chapters in political economy

American Empire and the Political Economy of Global Finance (International Political Economy)

Translating Empire: Emulation and the Origins of Political Economy

Authority in the Global Political Economy (International Political Economy)

The Theory of Political Economy

The Political Economy of Turkey

The Political Economy of Globalization

The political economy of stalinism

The Political Economy of Terrorism

The political economy of Stalinism

The political economy of health

The Theory of Political Economy

The political economy of theocracy

Models of Political Economy

Encyclopedia of political economy,

Classical Political Economy and British Policy in India (Cambridge South Asian Studies)

The Elements Of Political Economy

The Political Economy of Scotland

The Political Economy of Growth

The Political Economy of Sport

The Political Economy of Privatization

The Political Economy of Craft Production: Crafting Empire in South India, c. 1350-1650

Crafting the Nation in Colonial India

Craft Of Political Research

The Knowledge Economy in India

The Political Economy of Slavery: Studies in the Economy and Society of the Slave South

India: macroeconomics and political economy, 1964-1991

North and South in the World Political Economy

The Political Economy of Work (Routledge Frontiers of Political Economy)

Peasant History in South India

The Political Economy of Commerce: Southern India 1500-1650 (Cambridge South Asian Studies)

Chapters in political economy

American Empire and the Political Economy of Global Finance (International Political Economy)

Translating Empire: Emulation and the Origins of Political Economy

Authority in the Global Political Economy (International Political Economy)

The Theory of Political Economy

The Political Economy of Turkey

The Political Economy of Globalization

The political economy of stalinism

The Political Economy of Terrorism

The political economy of Stalinism

The political economy of health

The Theory of Political Economy

The political economy of theocracy

Models of Political Economy

Encyclopedia of political economy,

Classical Political Economy and British Policy in India (Cambridge South Asian Studies)

The Elements Of Political Economy

The Political Economy of Scotland

The Political Economy of Growth

The Political Economy of Sport

The Political Economy of Privatization

Recommend Documents